IDENTlFICATION AND CHARACTERIZATION OF DOWNSTREAM SIGNALING PARTNERS OF THE ElVDOTHELIAL CELL-SPECIFIC RECEPTOR , TEmIE-2

Nina Jones

A thesis submitted in conformity with the requirements for the degree of Ihxtor of Philosophy Department of Medical Biophysics University of Toronto

O Copyright by Nina Jones (2000) National Library BiMiothèque nationale du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 335 Wellington Street 395, rue Wdlim OttawaON K1A ON4 OttiwaON K1AW canada CaMde

The author has granted a non- L'auteur a accordé une licence non exclusive Licence aliowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distriiute or seli reproduire, prêter, distri'buer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/nlm, de reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or othedse de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. Identification and Characterization of Downstrem Signaling Partners of the Endothelial Cell-Specific . Temie-2. Doctor of Philosophy, 2000 Nina Jones Department of Medical Biophysics University of Toronto

ABSTRACT

Tek/Tie-2 is an endothelial cell-specific receptor tyrosine kinase that plüys a pivotai role in physiologicai and pathological angiogenesis. A family of ligands known as the ringiopoietins has recently been shown to signal through Tek although little is known about the molecules that are involved in Tek-mediated signal transduction pathways. The goal of this study was to determine exactly how Tek is eliciting its signais in endothelid cells through the identification of substrates and docking molecules that bind the activated receptor.

In order to circumvent the requirement for a ligand to specifically activate the receptor, we have used the yeast two-hybrid system to identify molecules that interact with phosphorylated Tek. Using this approach, we were able to demonstrate that a number of signaling molecules could associate with Tek, including the adaptor molecules

Grb2. Grb7 and Grbl4, the tyrosine phosphatase Shp2, the p85 regulatory subunit of phosphatidylinositol 3' kinase and one novel docking known as Dok-R.

Mapping of the binding sites of these molecules on Tek has revealed the presence of a rnulti-functional docking site on Tek that is required for binding of Grb2 and Grb7 and for tyrosine phosphorylation of Grb7 and p85. Activation of Tek by -1

II results in tyrosine phosphorylation of p85 as well as stimulation of endothelial ce11 migration and survival pathways that are partly dependent on phosphatidylinositol 3' kinase activity. These findings support the role of Tek signaling in sprouting angiogenesis and endothelid ceIl survival.

Dok-R contains numerous sites of tyrosine phosphorylation that mediate binding to RasGAP and Nck. Interestingly, overexpression of Dok-R downstream of the epidermal can attenuate mitogen activated activation and Dokr maps to a locus that is rearranged in a number of lymphoproliferative disorders.

These results suggest that Dok-R may be required for negative regdation of signal transduction.

Further characterization of these downstream signaling molecules should prove useful in elucidating the role of Tek within the endothelial ce11 Iineage. and this will ultimately contribute to our understanding of vascular development in both the embryo and the adult, as well as in pathological situations such as tumour-mediated angiogenesis. Acknowledgements First and foremost, 1 would like to thank rny supervisor Dr. Dan Dumont for the support, friendship and direction he has given me throughout my time in the Department of Medical Biophysics. From countless hours of discussion, Dan's words of advice and encouragement will long be remembered both in science and in life. 1 would also like to thank Dr. Mike Rauth, Dr. Jane McGlade and Dr. C.C. Hui, for their insightful critiques and guidance provided as members of my supervisory cornmittee. 1 am indebted to several past and present members of the Dumont lab for their technical expertise and assistance. Dr. Qiurong Liu, Dr. Li-Fong Seet, Renu Sarao, Martin Bissessar, Olga Agah, Stephen Cho and Stephen Chen. These thanks are also extended to Denis Bouchard of the AMGEN Institute and 1 am grateful to Sue Farinaccio for her friendship and invaluable administrative expertise.

The stresses of graduate student Iife were made much more bearable and even entertaining by the friendship, camaraderie and scientific discussion provided by feIlow graduate students. 1 would especially like to thank my friends Zubin Master and Stanley Liu for this. 1 am also grateful to my brother Jamie Jones (Moss) for his companionship both inside and outside the laboratory.

Throughout my life, my family has been a source of unconditional love and support and 1 thank my parents and siblings for their belief in me and my seemingly never ending scholastic career.

Most importantly, 1 would fike to thank Todd Porter for encouraging me to pursue this opportunity and for his commitment to me throughout. Without Todd, keeping me Company on those long nights in the lab, his patience with me in times of frustration, and his pride in me at times of accomplisliment, this story would have had a different ending.

Nina Jones Table of Contents .. Abstract 11

Acknowledgements i v

Table of Contents v

List of Abbreviations and Symbots viii

Chapter 1 Introduction

Mechanisms of Blood Vessel Development - The Process of Vasculogenesis - The Process of Angiogenesis - Blood Vessel Maturation

Endothelial Cell-Specific Receptor Tyrosine Kinases (RTKs) - Structure of the Endothelial Cell-Specific Subgroup of RTKs - Endothelial Cetl-Specific RTKs Involved in Early Stages of Blood Vessel Developrnent - Endothelial Cell-Specific RTKs Involved in Late Stages of Blood Vessel Development

Endotheliai Cell-Specific Peptide Growth Factors - The VEGF Family of Ligands - The Angiopoietin Family of Ligands - Angiopoietin Related (ARPs)

Tek Has Multiple Roles in Developmental and Pathological Angiogenesis - Tek Signaling is Required for Normal Blood Vesse1 Development pg. 17 - Tek Signaling is Involved in Turnour-Associated Angiogenesis Pg- 19 - The and Blood Vesse1 Stability pg. 21

Elements of Signal Transduction Pathways Mediated by RTKs pg. 2 1 1 S. 1 Modular Protein Interaction Domains Pg- 22 - The SH2 Domain pg. 33 - The PTE3 Domain Pg- 24 - PH and SH3 Domains pg. 25 1 -5.2 Classes of Cytoplasmic Signaling Molecules pg. 26 - Adaptor Molecules Pg- 26 - Docking Molecules Pg. 27 - pg. 29

1 S.3 Intracellular signaling pathways pg. 30

1 6 Signal Transduction Pathways Controlled by Tek

1.7 Rationale

1-23 Thesis Objectives and Organization pg. 36

1.9 Attributions pg. 37

Chapter 2 Identification of Teki'ïie-2 Binding Partnets: Binding to a Multi- Functional Docking Site Mediates Cell Survival and Migration.

2.1 Abstract pg. 40 2.2 Introduction pg 41 2.3 Materials and Methods Pg- 44 2.4 ResuIts pg. 49 2.5 Discussion pg. 69

Chapter 3 The Te k/Tie-2 Receptor Signals Through a Novel Dok-Related Docking Protein, Dok-R.

3.1 Abstract 3.2 Introduction 3.3 Materials and Methods 3.4 ResuIts 3.5 Discussion

Chapter 4 Assignment of the Dokr to Mouse 14D2-D3by Fluorescence in situ Hybridization.

4.1 Brief Mapping Report pg. 110 Chapter 5 Recruitment of Dok-R to the EGF Receptor Through its PTB Domain is Required for Attenuation of Erk MAP Kinase Activation.

5.1 Absuact pg. 115 5.2 Materials and Methods pg. 116 5.3 Results and Discussion pg. 117

Chapter 6 Discussion and Concluding Remarks

6.1 Summary of the Work pg. 136 6.2 Angiopoietin- 1 Induces Endothelia1 Cell Survival and Migration pg. 136 6.3 Tek Contains a Multisubstrate Docking Site that may be Regulated by Phosphatases pg. 142 6.4 The Unique Role of Doçking Molecules in Signal Transduction Pathways pg. 145 6.5 Conciuding Remarks pg. 15 1

Bibliography pg. 152

vii List of Abbreviations and Symbols

OC degree(s) Celsius

CLE microgram

PI micro1 iter

Pm micrometer

Ang angiopoietin

ARP angiopoietin related protein

ATP adenosine 5' -triphosphate bp base pairs

BSA bovine serum dbumin cDNA complementary DNA cm centimeter

CO- carbon dioxide

DAPI 4.6-diamidino-2-phen y I indoLedihydrochloride

DMEM Dulbecco's modified Eagle's medium

DNA deoxyribonucleic acid

Dok downstream of tyrosine kinases

DOKL ~62""'-likeprotein

Dok-R downstream of tyrosine kinases-related

ECGF endotheliai ce11 growth factor

E. coli Escherichia coli

EDTA ethylenediaminetetraacetic acid

EGF epidermal growth factor

viii EGFR epidermal growth factor receptor

EGTA ethylene glycol-bis (P-aminoethyl ether) N, N, N' , N' -tetraacetic acid

Erk2 extracellular signal-regulated kinase 2

FAK focal adhesion kinase

FBS fetal bovine serum

FIS H fluorescence in situ hybridization

FGF fibroblast growth factor

FRS2 FGF receptor substrate 2

Gab Grb2-associated binder

GDP guanosine 5'-diphosphate

Grb growth factor receptor bound protein

GST glutathione S-

GTP guanosine S'-triphosphate h hour

HA hemagglutinin

HCPTPA human cellular protein tyrosine phosphatase A

HEK human ernbryonic kidney

HEPES N-(2-hydroxyethyl) piperazine-N'-(2-ethanesulfonic acid)

HRP horseradish peroxidase

HUVEC human umbilical vein endothelial ce11

IC intracellular

IgG immunoglobulin G

IRS insu1 in receptor substrate kb kilobase pairs

Ki dissociation rate constant kDa kiloddton

M molar

MAPK mitogen activated protein kinase

MBD c-Met binding domain

MBP myelin basic protein mg milligram

MgCl: magnesium chloride min minute ml milliliter mM millimolar

NaCI sodium chloride

NaOH sodium hydroxide

"& nanogram

NGF nerve growth factor nM nanomolar

PBS phosphate buffered saline

PCR polymerase chain reaction

PDGF platelet-derived growth factor

PDK phosphoinositide-dependent kinase

PECAM platelet and endothelial ce11 adhesion rnolecule

PH pleckstrin homolo,~ P 1 phosphatidy linositol

PKB protein kinase B flGF placenta growth factor

PRR proline rich region

PTB phosphotyrosine binding

PTP protein tyrosine phosphatase

PY phosphotyrosine

RACE rapid amplification of cDNA ends

RasGAP p2 1 Ru.r GTPase-activating protein

RNA ribonucleic acid

RTK receptor tyrosine kinase

SDS sodium dodec y l sulphate

SDS-PAGE sodium dodecyl sulphate-polyacrylamide gel electrophoresis

SH2 src homology 2

SH3 src homology 3

Shc Src homolog and collagen homolog

Shp2 SH2 domain-containing protein tyrosine phosphatase 2

SOS Son of Sevenless

STAT signal transducer and activator of transcription

TBST Tris buffered saline with Tween 20

Tek/Tie-2 tunica interna endothelial ce11 kinase

Tienie- 1 tyrosine kinase wiih irnmunoglobulin-like loops and epidemal growth

factor homology domains

xi VEGF vascular endothelial growth factor

VEGFR vascular endothelial growth factor receptor

VE-PTP vascular endothelial protein tyrosine phosphatase

WCL whole ceIl iysate

Y tyrosine

xii CHAPTER 1

INTRODUCTION. 1.1 Mechanisms of Blood Vesse1 Deveiopment

The cardiovascular system, which is composed of the hem, blood vessels and

bIood. functions to supply essentiai oxygen and nutnents to body tissues and to eliminate

waste substances from these tissues. During embryonic development, the cardiovascular

system is formed through a number of complex processes that involve the coordinated

interactions between distinct ce11 lineages. Endothelial cells are extremely flattened epitheiial cells that form a monolayer to line the lumen of blood vessels and provide an interface between the bloodstream and the vesse1 wall. Mature endothelial cells from a particular tissue or organ acquire diverse specialized functional and ultrastructurat characteristics while also displaying heterogeneity in their expression of ce11 surface- associated proteins (Auerbach et al., 1985; Risau, 1995). In mature blood vessels, this population of cells is relatively quiescent compared to many other ce11 types in the body because the turnover of these cells is very low (Engerman et al.. 1967). In contrast. during embryonic development, endothelial cell proliferation is high to allow for the expansion of the vasculature. The regulated proliferation, migration and differentiation of endothelial celIs during this period can be divided into two discrete proccsses known as vasculogenesis and angiogenesis (Risau et ul., 1988).

The Process of Vasculugenesis

Vasculogenesis takes place during embryonic development and it is the earliest formation of blood vessels in the embryo. During vasculogenesis, mesodermally-derived endothelial ce11 precursors known as angioblasts differentiate and assemble into discrete blood vessels in sitir to form a primitive vascular network consisting of relatively uniformly sized endothelial channels (Risau and Flamme, 1995) (Figure 1.1). This process is required for vascularization of the major vascular structures of the early ernbryo that are of endodermal origin, including the spleen, the heart endocardiurn and the dorsal aorta as well as the extra-embryonic yolk sac (Coffin and Poole, 1988;

Pardanaud et al., 1987). It has recently been established that vasculogenesis may not be restricted to embryogenesis and that this process could also contribute to postnatal neovascularizarion- Circulating endothelial progenitor cells have ken isolated (Asahara et cd., 1997: Hatzopoulos et al., 1998; Shi et al., 1998) and these angioblast-Iike cells cm home to sites of vascularization in the adult with the capacity to differentiate into endothelial cells in situ, consistent with vasculogenesis (Takahashi et al., 1999).

The P rocess of Angiogenesis

Once a primq vascular plexus is forrned, it is subsequently remodeled through the process of angiogenesis whereby new capillaries arise from pre-existing larger vessels to give rise to a more cornplex vascular network with a hierarchy of both large and small vessels (Figure 1.1 ). At least three distinct types of angiogenesis have been described and they are classed as either sprouting or non-sprouting angiogenesis (reviewed in

Risau, 1997; Wilting and Christ. 1996) (Figure 1.2). During sprouting angiogenesis, new vessels forrn by budding, sprouting and extending from extant larger vessels in response to an angiogenic stimulus. Endothelial cells must first invade the surrounding basement membrane and they begin to proliferate at the base of the developing sprout.

Differentiation of the endothelial cells behind the leading edge causes the cells to change shape and adhere to one another to forrn a lumen. Adjacent tubes then arch and MESWERML ENDOTHECIAL PRECURSORS CELL ASSEMBLY

PRIMARY COMPLEX CAPILCARY VASCULAR PLEXUS NETWORK

Vesse1 Maturation

PDGF-B SMOOTH MUSCLE CELLS MATURE L0 OR PERICYïES SURROUNO VESSEL ENOOT'HELIAL CELLS

Figure 1.1: Mode1 of vascular development. In vascuIogenesis, endothelial cells differentiate and assemble in situ into primitive tubes. This primary plexus is then remodeled through the processes of angiogenesis where new vessels arise from pre- existing vessels. Support cells are then recruited to the endothelial tubes (shown in cross-section) to stabilize the vessel. Adapted from Beck and D' Amore, 1997. Figure 1.2. Schernatic illustration of sprouting and non-sprouting angiogenesis. (A) Angiobiasts aggregate to form lumenized vessels. (B) Endothelial cells form sprouts which anastomose into capillary loops. (C) Intussusceptive growth where transluminal pillars create vascular loops. (D) Fusion of blood vessels. (E) Intercalated growth of vessels to increase diameter and length. Adapted from Wilting and Christ, 1996. eventually fuse to form a newly perfused loop. Sprouting angiogenesis is the principal mode of vascularization during organogenesis of the brain which does not contain angioblasts (Risau. 1997; Stewart and Wiley, 198 1) as well as the kidney (Sariola et al..

1983) and the intersornitic arteries (Coffin and Poole, 1988). Angiogenic sprouting is also required in the adult for the continuous remodeling of the female reproductive system as well as in wound healing (Sholley et al., 1984) and in pathological processes such as tumour growth (Ausprunk and Folkman. 1977; Folkman, 1995).

The first type of non-sprouting angiogenesis is referred to as intussusceptive microvascular growth and it involves the splitting of large pre-existing vessels into two smaller capillaries that grow separately (Patan et al.. 1996). In this process, endothelial cells sprout into the lumen of the vessel to produce transcapillary pillars which fuse with the opposite side of the vessel to create a vascular loop. In the second type of non- sprouting angiogenesis, endothelial cells are intercalated into an existing vessel to increase both the diameter and length of the vessel and this intercalated growth of blood vessels is important for re-endothelialization of vessels following injury (Carmeliet et al.,

1998: Risau, 1997). Non-sprouting angiogenesis predominates in the Iung which is initially vascularized by vasculogenesis (Pardanaud et al., 1989), and concurrent sprouting and non-sprouting angiogenesis also occur in the heart and yolk sac vasculature

(Risau. 1997). Blood Vessel Mahrrdon

Upon establishment of the primary vascular network, peciendothelial support cells are recruited to the nascent vessels to surround the endothelial tubes (reviewed in

D'Amore, 1992) (Figure 1.1). These supporting stroma1 cells belong to a different lineüge than endothelial cells and they are referred to as microvascular pericytes when they surround small capillaries, smooth muscle celIs when they surround larger vessels and myocardiocytes when they support the heart. The support cells are gradually attracted to the vascular wall by growth factors such as platelet-hrived growth factor

(PDGF)-B which are derived from the endothelium itself and this leads to a progressive covering of the endothelial plexus (Benjamin et al.. 1998; Leveen et al., 1994: Lindahl er cil.. 1997). Interactions between the endothelial cells and these support cells are thought to lead to inhibition of endothelial ce11 migration and proliferation and to stabilization of the vessel (Antonelli-Orlidge et al., 1989: Orlidge and D'Amore. 1987; Sato and Rifkin.

1989). During the final stages of blood vessel remodeling, unprotected vessels regress and the pericyte or smooth muscle ceII-covered vessels become stabilized and less prone to regression.

1.2 Endothelial Cell-Specific Receptor Tyrosine Kinases (RTKs)

Structure of the Endothelial Cell-Specific Subgroup of RTKs

Cellular events such as growth and differentiation are often controlled by molecular signal transduction pathways that are mediated by ce11 surface growth factor receptors known as receptor tyrosine kinases (RTKs). RTKs are membrane-spanning proteins that consist of a hydrophobic transmembrane region separating a complex extracellular ligand binding domain from an intracellular domain which contains a juxtamembrane region, a catalytic tyrosine kinase domain and a carboxy-terminal tail. A number of these receptors have been identified whose expression is almost exclusively restricted to cells of the endothelial lineage (reviewed in Partanen and Dumont, 1999;

Shibuya et al., 1999; Taipale er al., 1999). This subgroup of RTKs has been placed into two distinct subfamilies according to similarities in their prirnary amino acid sequences

(Dumont et al.. 1993; Mustonen and Aiitalo, 1995). The first subfamily contains the high affi nity xascular endothelial growth factor (VEGF) sceptors (VEGFRs) denoted

VEGFR- 1 (Flt- 1 ). VEGFR-2 ( KDiUFlk- 1 ) and VEGFR-3 (Flt-4) (Figure 1 -3). The extracel lular domains of the VEGFRs are characterized by seven imrnunoglobulin-l i ke motifs that are forrned by disulphide links between paired cysteine residues. VEGFR-3 is unique in that it is proteolytically cleaved into two disulfide-linked polypeptides in the extracellular domain (Pajusola et al., 1993). In the tyrosine kinase domains of these receptors, the ATP binding lobe of the is separated from the phosphotransferase activation lobe by a srnaII stretch of amino acids commonly referred to as a kinase insert.

Neuropilin-l has also been identified as a receptor for VEGF although this receptor is unique from the VEGFRs in that it does not appear to have tyrosine kinase activity and it is expressed abundantty by both endothelial and non-endothelial cells such as neuronal and tumour ceils (Soker et al., 1998).

Tie (Tie- 1 ) (tyrosine kinase with ~mmunoglobulin-like loops and epidemal growth factor homology domains) and Tek (Tie-2) (iunica interna endothelial ce11 kinase) comprise the second subfamiiy of endothelial cell-specific receptors and they are Angl

Ang2

VEGF-B V EGF-C Ang3

PlG F VEGF (A) VEGF-D Ang4

VEGFR-1 N RP- 1 V EG FR-2 V EG FR-3 Tiemie- 1 Temie-2

Migration, Proliferation, Permeability, Survival Migration, Survival

Figure 1.3: Endothel ial celf -specific receptor tyrosine kinases and their ligands. The structures of the receptors are shown schematically and the major protein domains are indicated. The loops represent immunoglobulin homology domains, the grey ellipses are epidermal growth factor-like reps, the grey boxes represent fibronectin type III-like motifs and the black boxes are the tyrosine kinase catalytic domains. Activation of the receptors Ieads to different cellular responses depending on the ligand involved. NRP- 1, -1. Adapted from Ferrara and Alitalo, 1999. stmcturally divergent from the VEGFRs (Dumont et al., 1992; Dumont et al., 1993;

Honta et al., 1992: Iwama et al., 1993; Maisonpierre et al., 1993; Partanen et al., 1990;

Partanen et al., 1992; Sato et al., 1993; Schnurch and Risau, 1993; Ziegler et al., 1993).

The extracellular domains of Tie and Tek are quite cornplex and they consist of three different structural motifs that are not usuaIIy found colIectively within a single RTK.

The extracellular domains contain two immunoglobulin-Iike loops that are separated by three tandem epiderrnal growth factor (EGF)-like cysteine repeats and followed by three fibronectin type III-like motifs. Sirnilar to the VEGFRs, the tyrosine kinase domains of both Tie and Tek are intempted by a short stretch of amino acids. These two subfarnilies of RTKs dispIay distinct but overlapping expression patterns in the developing embryo. suggesting that the signaling pathways that are mediated by these receptors play unique roles in vascular deveIopment and within the endothelial ce11 lineage (Dumont et al.,

1995).

EndotheLial CeLL-Specific RTKs Involved in Eariy Stages of Blood Vessel Development

The importance of these RTKs has been underscored in mouse molecular genetic experiments where it has been demonstrated that signaling by each of these receptors is absolutely required for a discrete phase of early cardiovascular development. Consistent with their expression profiles, VEGFR-2 and VEGFR-I appear to be involved in early- and late-stage vasculogenesis, respectively (Figure 1.4). VEGFR-2-deficient embryos display defects in the establishment of both the endothelial and hematopoietic lineages

(Shalaby et ul., 1995) and this receptor has been shown to play an additional role in both primitive and definitive hematopoiesis (Shalaby et al., 1997). In VEGFR- 1-deficient VEGFR-2 VEGF

VEGFR-1 VEGFR-2 VEGFRS VEGF SPROUTING, MGRATION, ORGANRATlON TEK ANGl

'JUVENILE' VASCULAR SYSTEM

TIE + TEK REMODELNG AND REGRESSIW ANGl SUPPORT CELL RECRUITMENT ANGZ

MATURE VASCULAR SYSTEM

VEGF Tl€ SURVIVAL, MAINTENANCE

Figure 1.4: Regulation of vascuiar development by RTKs expressed on endothelial cells and their ligands. Additional RTKs and ligands not mentioned in the text have been omitted from this schematic. Adapted from Risau, 1997. embryos, the endothelial lineage is developed however the endothelial cells do not arrange into an organized vascular plexus and there appears to be an overgrowth of endothelial cells within the vessel lumen (Fong et al., 1995). Interestingly. the tyrosine kinase activity of VEGFR-1 has been shown to be dispensable for embryonic vascular development and given the greater affinity of VEGFR-1 for VEGF than VEGFR-2,

VEGFR- 1 may function to bind significant amounts of VEGF in vivo to negatively regulate endothelial growth (Hiratsuka et al., 1998). Similarly, alterations in Neuropilin-

1 expression result in severe abnorrnalities in the development of the cardiovascular system as well as the peripheral nervous systern although the precise function of VEGF signaling through Neuropilin- 1 is unknown (Kitsukawa et al., 1995: Kitsukawa et al..

1997).

Enào thelial Cell-Specific RTKs lnvolved in Lute Stages of Blood Vessel Development

In contrat to the roles of VEGFR-1 and -2 in vascuiogenesis, VEGFR-3. Tie and

Tek function later in angiogenesis during vascuIar remodeling and vessel maturation

(Figure 1.4). VEGFR-3-deficient mice manifest defects in the organization of large vessels surrounding the heart (Dumont et al., 1998) and expression of VEGFR-3 becomes largely restricted to lymphatic endothelium later in deveIopment suggesting an additional role for this receptor in the maintenance of lymphatic vrisculature (Kaipainen et nt.,

1995). Tie-deficient mice display edema and localized hemorrhaging resulting from compromised vascular integrity and they succumb later in the embryo or as neonates

(Puri et al., 1995; Sato et al., 1995). Lady, Tek-deficient mice do not undergo sufficient sprouting and remodeling of the primary capillary plexus that resul ts in incornplete development of the heart and head regions (Dumont et al., 1994; Sato et al.. 1995).

There is also a dramatic reduction in the number of endothelial cells in these mice

(Dumont et al., 1994) and embryos Iacking Tek exhibit defects in in vitro assays of

definitive hematopoiesis (Takakura et al.. 1998; Puri et al.. 1999). In contrast to the

findings with VEGFR- 1, the tyrosine kinase activity of Tek is critical for embryonic

angiogenesis since expression of a catalytically inactive receptor in transgenic mice

resuIts in embryonic lethality with similar defects as those seen in mice lacking Tek

(Dumont et al., 1994). Taken together, the overlapping pleiotropic defects seen in these mice demonstrate that the coordinated regulation of a multitude of signaling pathways exists during early vascular development.

1.2 Endothelial Cell-Specific Peptide Growth Factors

The VEGF Family of Ligands

VEGF was the first selective angiogenic growth factor purified and it was simultaneously identified based on its ability to induce transient vascular leakage. thus it was also named vascular permeability factor (Senger et al., 1983). To date. five VEGF- type ligands have been identified, including VEGF-A (herein referred to as VEGF),

VEGF-B, VEGF-C, VEGF-D and aacenta growth factor (PIGF), and they are encoded

by separate (reviewed in Neufeld et al., 1999; Ferrara, 1999). As illustrated in

Figure 1.3, the five VEGF ligands bind and activate different subsets of the VEGFRs.

Notably, distinct isoforms of VEGF and VEGF-B exist as a result of alternative splicing and Neuropilin- l functions as an isoform-specific receptor for VEGF (Soker et al., 1998).

VEGF has a unique combination of properties, allowing endothelial cells to proliferate, migrate and assemble into tubes and it also functions as a potent survival factor (Alon et ai., 1995). VEGF has been shown to be essential for the development of the vascular system. Loss of a single VEGF allele causes embryonic IethaIity in mice as a resuit of numerous defects in early blood vessel formation (Camieliet et al., 1996; Ferrara et ai.,

1996). Remarkably, even selective inactivation of particular VEGF isofonns is sufficient to disrupt normal cardiovascular developrnent (Carmeliet et al., 1999). This notwithstanding, a more severe phenotype is seen in rnice lacking VEGFR-2 than in those lacking VEGF, indicating the importance of additional VEGF-li ke ligands for this receptor such as VEGF-C (Joukov et al., 1996). These findings indicate tight dose- dependent control of blood vessel development by the VEGF family of growth factors.

The A ngiopoietin Family of ligands

A second family of growth factors specific for the vascular endothelium has recently begun to emerge and these unique ligands are known as the angiopoietins (Davis et cil., 1996; Maisonpierre et al., 1997; Mezquita er al., 1999: Valenzuela er al., 1999).

The angiopoietins isolated to date appear to bind exclusively to the Tek receptor and the lisand for the closely related Tie receptor remains to be identified. This ligand family is characterized by the presence of an amino-terminal hydrophobie secretory signal sequence. a central coi led-coi1 region and a carboxy-terminal fibrinogen-li ke domain wi th three closely spaced cysteine residues. Angiopoietin- 1 (Ang 1 ) was first isolated using a novel secretion-trap expression cloning strategy and it functions as an agonistic or activating ligand for Tek in endothelid cells (Davis et al., 19%). Angiopoietin-2 (Ang2) was subsequently cloned and shown to be approximately 60% identical to Ang 1 with a simiiar affinity for Tek (Maisonpierre et al., 1997). Ang2 is expressed in endothelial cells while Ang 1 is expressed in the surrounding smooth muscle cells (Davis et al.. 1996:

Maisonpierre et al., 1997). Interestingly, Ang2 appears to act as a cornpetitive antagonist for Tek in a ce11 type-specific manner as it blocks the ability of Ang 1 to activate Tek in endothelial cells while it activates Tek ectopically expressed in fibroblasts (Maisonpierre et al., 1997) and in hematopoietic precursor cells (Sato et al., 1998). The opposing actions of these structurally similar ligands have been ascribed to the receptor-binding fibrinogen-Iike domains while the coiled-coi1 domains mediate distinct multimerization patterns of the ligands in vitro (Procopio er al., 1999; Valenzuela et al., 1999). Since ligand binding alone is not sufficient to activate Tek in endothelial cells, it is likely that these cells have some unique components that allow functional discrimination between the two angiopoietins.

AdditionaI members of the angiopoietin family have been isolated using homology-based cloning approaches. Angiopoietin-3B (Ang2B) is an avian isoforrn of

An@ with a partially truncated amino-terminal coiled-coi1 domain that exists as result of alternative splicing of a 5' intron (Mezquita et al., 1999). Interestingly, the expression profiles of An@ and Ang2B are distinct although the functionaI significance of this is unknown (Mezquita et al., 1999). Angiopoietin-3 (Ang3) is a context-dependent antagonist for Tek similar to An@ while Angiopoietin-4 (Ang4) functions as an agonist for Tek similar to Angl (Valenzuela et al., 1999). Mouse Ang3 and human Ang4 are actually interspecies orthologs that are more divergent than the mouse and human counterparts of Ang 1 and Ang2 and this structural divergence is associated with dramatic differences in their function and expression (Valenzuela et al., 1999). The distinct but overlapping expression patterns of the angiopoietins suggests that each may possess a unique function that is required for the spatial and temporal regutation of vascular development-

A ngiopoieîin Related Proteins (ARPs)

The angiopoietins are members of the fibrinogen superfamily that also includes fibrinogens and lectins and a fourth member referred to as the Angiopoietin Related

-Proteins (ARPs) has recently been identified (Kim et al., 1999a; Kim et al., 1999b).

These widely expressed proteins have been termed ARPs since they are closer to the angiopoietin family than the other two families and also because they contain amino- terminal coiled-coi1 domains. ARP2 is approximately 60% identical to ARPl although the ARPs are only about 35% identical to Angl-4 (Kim et al., 1999b). Somewhat confusingly, ARPl was initially named Ang3 (Kim et al., 1999a) although ARPl is different than the Ang3 reported simultaneously by Valenzuela et al. (1999). Based on amino acid similarity, Ang3 and Ang4 were closer to Angl and Ang2 and it was subsequently renamed ARP1 (Kim et al., 1999b). ARP1 and ARP2 do not bind Tie or

Tek i)z vitro although they do induce endotheliai cell sprouting and it remains to be determined whether a novel yet unidentified receptor for these ligands exists (Kim et al.,

1999b). 1.4 Tek Has Multiple Roles in Developmental and Pathological Angiogenesis

Tek Signalhg is Required for NodBIood Vessel Development

Gene targeting studies in mice have revealed that signal transduction pathways mediated by Tek are indispensable for normal embryonic blood vessel development.

Disruption of the Tek agonistic ligand, Ang 1, results in embryonic lethality with defects in angiogenesis that are strikingly similar to those seen upon disruption of Tek (Dumont er cil.. 1994: Sato et al,, 1995; Suri et al., 1996). Interestingly however, the defects observed in these mice are less severe than those observed in mice lacking Tek, implicating the additional angiopoietins in Tek function. Similarly. transgenic overexpression of Ang2 in endothelial cells results in vascular defects that resemble those seen in the absence of Ang 1 or Tek (Maisonpierre et al., 1997), dernonstrating that Ang2 can potentially regulate Ang 1 in vivo by antagonizing the effects of Ang 1 on Tek.

Embryos lacking Tek signaling pathways display a generai underdevelopment of the vasculature with impaired endocardial development and myocardial trabeculation that has been attributed to defects in both sprouting angiogenesis and intussusceptive microvascular growth (Patan, 1998). Close ultrastmctura1 analysis of the vessek in these embryos has revealed a paucity of pericytes and srnooth muscle cells at sites of deficient

vessel branching and in the heart region, the endocardium and the underlying

myocardium are poorly associated (Dumont et al., 1994; Patan. 1998; Sato et al., 1995:

Suri et al., 1996). In contrat, mice overexpressing Ang 1 in the skin exhibit an increase

in the diameter of blood vessels (Suri et al.. 1998) and these vessels have subsequently been shown to be resistant to leakage induced by inflammatory stimuli, even in the presence of excess VEGF (Thurston et al.. 1999).

From these findings, it has been hypothesized that Tek signaling may be required to stabilize mature blood vessels by stimulating the release of chemoattractant growth factors from endothelial cells to facilitate recruitment and tight association with adjacent periendothelial cells (Patan, 1998; Sato et al., 1995; Suri et al., 1996; Vikkula et al.,

1996). However, recent mosaic experiments have clearly demonstrated that endothelial celIs lacking Tek can still contribute normally to early embryonic blood vessels, including those participating in angiogenic sprouting (Puri et al.. 1999). Notably however, these Tek-deficient cells were virtually excluded from the microvasculature of late embryonic and adult tissues (Puri et al., 1999). These observations support an earlier report of a progressive loss of endothelial ceIls in late-stage Tek-nuIl embryos where it was suggested that these cells may be undergoing an active ce11 death process (Dumont et rd., 1994). These findings strongly suggest that Tek is not directly necessary for early angiogenic sprout formation and perivascular cell recruitment and the primary function of

Tek is likely to transduce an ongoing survival stimulus to endothelial cells. This apparent failure to recruit periendothelial cells may therefore be an actual loss of contact between the endothelial and smooth muscle ceIl layers as a consequence of endothelial ceIl apoptosis.

The physiological importance of Tek signaling fias been underscored by the observation that Tek is expressed at relatively high levels in the quiescent vasculature of a number of adult tissues as well as in healing skin wounds (Wong et al., 1997; Yuan et

cd., I999). Interestingly, Tek isolated from these numerous different tissues was

phosphorylated, suggesting that the receptor is active in quiescent endothelium. Gain-of-

function mutations at the Tek locus have also been identified in some families with

inheri ted venous malformations (Calvert et al.. 1999; Vikkula et al., 1996). Although the

precise consequences of these mutations in signal transduction pathways rnediated by

Tek have not been defined, it has recently been show that one of these mutant forrns of

Tek can uniquely activate signal eansducer and activator of transcription 1 (STAT 1 )

(Korpelainen et al., 1999). In summary, these experiments indicate that Tek and the angiopoietins are required throughout embryonic development and in the adult and they may influence at least two different signal transduction cascades that can effect changes in angiogenic remodeling and vesse1 maintenance.

Tek Signaling is Involved in Tumour-Associated Angiogenesis

The development of solid tumours beyond a certain size depends on their ability to induce the growth of new b1ood vessets and the production of angiogenic factors by the tumour cells is essential for neovascularization (Folkman, 197 1). In addition to the critical role of VEGF in regulating normal embryonic vascular development, VEGF is also a prime contributor to turnour-associated angiogenesis. VEGF is a pro-angiogenic factor released from most tumour cells and it functions in a paracrine manner to bind and activate VEGFRs expressed on the surface of tumour endothelial cells. resulting in the formation of new vessel-like structures (Ferrara, 1999). Interestingly however, recent studies are beginning to suggest that Tek signaling pathways are also involved in tumour- mediated angiogenesis.

Tek is highly upregulated in the endothelium of most malignant breast cancers. particularly in intense areas of neovascularization referred to as vascular hot-spots found at the periphery of invasive carcinomas (Peters et al., t 998). The expression of Tek in tumour vasculature suggests that an aitemative to the VEGFR pathways exists for tumour angiogenesis and inhibition of Tek signaling may provide a potential treatment for tumours that do not respond well to VEGF therapy (Millauer et al,, 1994; Millauer et al.,

1996). Recombinant soluble extracellular domains of Tek (ExTek and sTek) that function as dominant negatives by competing with the endogenous receptor for ligand binding and inhibiting receptor activation have been used as angiogenesis inhibitors (Lin et cd., 1997; Lin et al., 1998; Siemeister et al., 1999). When administered to mice bearing prirnary or established tumours, ExTek was shown to substantially inhibit tumour growth and metastases (Lin et al., 1997; Lin et al., 1998). Imponantly, ExTek did not affect the viability of cultured tumour cells, demonstrating that the effect was due to inhibition of tumour angiogenesis rather than direct tumour toxicity (Lin et al., 1997). In an unrelated study, human melanoma ceHs which produce Angl, VEGF and VEGF-C were engineered to express sTek and in nude mouse xenografts, tumours derived from these cells were smaller with a reduction in microvessel density when compared to control tumours (Siemeister et al., 1999). Together these findings demonstrate that Tek expression in tumour vasculature may serve as an important prognostic indicator and that interference with Tek signaling pathways may provide an effective anti-angiogenic strategy.

The A ngiopoietins and BIood Vessel Stabiliîy

With the discovery of the angiopoietins. a second mode1 to describe the mechanism of tumour angiogenesis has been proposed. Many turnours growing in the confines of a vascularized tissue do not initially grow as avascular masses and instead they can coopt existing blood vessels. These vessels subsequently regress prior to the initiation of angiogenesis at the tumour margin. Both vessel regression and angiogenic sprouting are associated with a progressive disengagement of endothelial cells from the surrounding support cells and Ang2 is upregulated in these vessels prior to the expression of VEGF in the adjacent tumour cells (Holash et al., 1999; Stratmann et al., 1998:

Zagzag et al., 1999). These data have suggested that Tek signaling may maintain the contacts between endothelial cells and surrounding support cells although the mechanism by which this occurs remains to be determined (Puri et al., 1999). This notwithstanding, tightly regulated expression of Ang2 may block the constitutively stabilizing function of

Ang 1 and this could explain the overlapping expression profiles of Ang 1 and Ang2 in particular regions of the devetoping embryo and adult that require precisely controlled vessel remodeling (Davis et 01.. 1996; Maisonpierre et al., 1997).

1.5 Elements of Signal Transduction Pathways Mediated by RTKs

Insight into the molecular signahg mechanisms that control vasculogenesis and angiogenesis has been provided by studies of the signal transduction pathways that underly the functional differences between the endothelial-specific RTKs. Signal transduction is initiated when a particuiar growth factor or ligand binds a receptor, resulting in receptor clustering at the cell surface (reviewed in van der Geer et al., 1994).

This ligand-mediated receptor oligornerization triggers the activation of the intrinsic tyrosine kinase activity of each receptor and the activated cytoplasmic domains then mutually transphosphorylate adjacent receptors within the oligomer.

Autophosphorylation of the receptors occurs at a specific set of tyrosine residues that cm serve as high affinity binding sites for numerous intracellular signaling molecules that specifically recognize phosphotyrosine. Once associated with a receptor. these intracellular signaling molecules may a1so serve as substrates for the autophosphorylated receptors and this allows for the recruitment of additional downstream effector molecules to the complex. The unique series of signaling moiecules that interacts with an individual activated tyrosine kinase contributes to the biological specificity of the receptor and these proteins serve to connect the receptor to downstream signal transduction cascades that ultimately control biological responses such as ce11 growth, differentiation and survival.

1S. 1 Modular Protein Interaction Domains

A regulated cascade of protein-protein interactions allows numerous signaling molecules to be recniited and assembled into an organized biochemicai network, leading to the activation of important signaling pathways within the cell. Associations between signaling molecules are mediated by small, distinct and independently folded regions of host proteins that recognize particular motifs on target proteins. Some binding modules recognize changes in phosphorylation States of other proteins and these types of inducible interactions allow relocalization of host proteins to particular subceIlular sites. In contrat, other interaction domains are not dependent on modifications of target proteins and these domains regulate the assembly of pre-formed multiprotein complexes within the cell. As a result. these conserved modular domains impart a particular function to the molecule in which they are found and this contributes to the specificity of signal transduction events.

Tire SH2 Domain

One such specialized protein module found in a variety of cytoplasmic signaling molecules is the Src brnology 2 (SH2) domain, first identified in the Src cytoplasmic tyrosine kinase (Sadowski et al., 1986). The SH2 domain binds with high affïnity directly to phosphotyrosine residues found on autophosphorylated RTKs and tyrosine phosphorylated intracelMar signaling molecules (Anderson et al., 1990; Koch et al.,

199 1: Moran et al., 1990). Binding of an SH2 domain to a tyrosine-phosphorylated protein depends on both the primary amino acid sequence of the SH2 domain itself as well as the sequence surrounding the phosphotyrosine in the target protein. SH2 domains have a well-conserved deep phosphotyrosine binding pocket (Eck et al., 1993; Pascal et al.. 1994: Waksman et al., 1993) that is followed by a more variable groove or a second binding pocket that associates with the residues carboxy-terminal to the phosphotyrosine

(the +I to +3 positions). One exception is the SAP SH2 dornain which has three binding pockets that associate with both the amino- and carboxy-terminal regions of the target as welI as the phosphotyrosine and engagement of any two of the three pockets provides sufficient binding energy (Li et al., 1999: Poy et al., 1999). The ability of SH2 dornains to discriminate between different phosphotyrosine-containing motifs provides a high

level of binding specificity in signal transduction (Marengere et al., 1994; Songyang et al.. 1993).

The PTB Domain

A second protein module that also mediates interactions with phosphotyrosine- containing motifs is known as the p_hospholyrosine bnding (PTB) domain.

Simultaneously identified by three independent groups, the PTB domain was initiaIly referred to as the phosphotyrosine interaction domain (Blaikie et al., 1994), the PT% domain (Kavanaugh and Williams, 1994) and the Shc and 'RS-1 NPXY (SAIN) binding domain (Gustafson et al., 1995). PTB domains can recognize and interact with phosphotyrosine residues contained within asparagine-proline-X-tyrosine (NPXY, in the single-letter amino acid code where X is any amino acid) motifs in target proteins

(Gustafson et al., 1995: Kavanaugh et al., 1995; van der Geer et al.. 1996). Interestingly, unlike the SH2 domain, the PTB domain is not strictly limited to binding phosphotyrosine-containing peptides. Numerous PTB domains cm bind their targets in a non-phosphotyrosine or non-NPXY motif dependent manner (Borg et al.. 1996: Chien et cal., 1998: Dho et al., 1998; Li er al., 1998) and some can also bind targets that do not contain tyrosine (Charest et al., 1996). These findings dernonstrate that PTB domains have evolved different requirements for the presence of phosphotyrosine in their target motifs and it suggests that the specificity of PTB domain binding may be more dependent upon tertiary structure than on a linear sequence. The overall sequence conservation among PTB domains is quite low although the three-dimensional structures are rernarkably similar. In contrast to SH2 domains, there are no strict conformational limits on the interaction between a PTB domain and its peptide ligand since the phosphotyrosine is not buned in a tight pwket (Eck et al.. 1996; Zhou et al.. 1995). This may explain how the structure of the PTE3 dornain can accommodate differences in target peptides.

PH and SH3 Domains

Additional protein-protein interaction motifs exist that are not dependent on phosphotyrosine. These domains include the gleckstrin homology (PH) and Src-

-homology 3 (SH3) domains. The PH domain is structurally related to the PTE! domain despite its sequence divergence (Macias et al., 1994) and it interacts primarily with phosphorylated inositol cornpounds found on the inner face of the plasma membrane

(Lemmon et al., 1996). htracellular signaling proteins containing PH domains appear to be localized close to the membrane in a signal-dependent manner. In contrast to the inducible. phosphorylation-dependent interactions of other protein modules. SH3 domains bind constitutively to short polyproline motifs in target proteins (Ren et al..

1993). allowing the assembly of pre-forrned protein complexes in the cytoplasm. The main focus of this discussion bas been on SH2. PTB, SH3 and PH domains although numerous other protein-protein interaction motifs exist, each with a unique function and binding specificity, and collectively these distinct domains contribute to the regulation of signal transduction cascades. 1.5.2 Classes of Cytoplasmic Signaiing Molecules

These conserved binding modules are often found in multiple copies and in association with other recognition modules in a single protein. Various combinations of protein-protein interaction domains allow a vast array of signaling proteins to be assembled with each unique signaling molecule having a precisel y defined function and an ability to bind multiple targets. Such domains may also be covalently associated with catalytic domains such as kinases and phosphatases. There are various classes of cytoplasmic signaling molecules that can associate with activated RTKs and collectively these proteins coordinate a diverse network of molecular interactions that connect activated receptors to downstream cellular events.

Adaptor Molecules

Adaptor molecules lack intrinsic catalytic activity and their primary function in signal transduction is to CO-localizespecific sets of proteins by mediating protein-protein interactions with other signaling molecules. These proteins typically contain SI32 domains as well as one or more additional protein interaction domains such as an SH3 or

PH domain. The SH2 domain functions to bind specific phosphotyrosine residues on activated RTKs while the additional domains simultaneously interact with effector molecules which are subsequently cotranslocated to the receptor. These bound effector molecules can serve as substrates for the activated RTK or they may possess enzymatic activity, and these proteins in turn propagate the signal. Growth factor gceptor -und protein 2 (Grb2), Nck and Crk are adaptor proteins that are composed entirely of SH2 and

SH3 domains while the Grb7 family of adaptor proteins, including GrblO and Grb14, consist of an amirio-terminal proline gich ~egion(PRR), a central PH domain and a carboxy-terminai SH2 domain (Figure 1 S).

Docking Molecules

An alternate strategy utilized by some receptors to coordinate multiple signal transduction pathways is the recruitment of a specific type of adaptor protein known as a docking or scaffolding molecuie. These molecuies bind the activated receptor and become phosphorylated on multiple tyrosine residues that serve as docking sites for the

SH2 domains of numerous intracellular signaling proteins. Many docking molecules contain more potential SH2 binding sites than the receptor itself. They therefore function as scaffolds to assemble signaling comptexes and they provide a rnechanism for signal amplification and diversity. The insulin Ieceptor substrate (IRS) family is the rnost well characterized group of docking molecules. A11 four members bear an amino-terminal PH domain. a central PTB domain and a carboxy-terminal tail rich in tyrosine and proline residues which provide binding sites for SH2 and SH3 domain-containing molecules. respectively (Figure 1.5). The PH and PTB domains act cooperatively to target the IRS proteins to the membrane where they become phosphorylated on tyrosine residues which links thern to SH2 domain-containing signaling molecules (Dhe-Paganon et al., 1999).

Tyrosine phosphorylated IRS proteins can then associate with numerous SH2 domain- containing signaling molecules, linking the to specific signal transduction pathways. A second group of docking molecules known as the Src ~omologand collagen hornolog (Shc) family is unique in that Shc proteins contain both an SH2 and a PTB domain as well as tyrosine phosphorylation sites (Pelicci et al., 1992) (Figure 1 S). Adaptor Molecules

1 Grb7 - PRR -- PH

Nck SH3 SH3 SH3

i Crk SH3 SH3 -

Docking Molecules

IRS- 1 PH *

Enzymes

RasGAP PH - GAP m

Figure 1.5: A schematic representation of selected cytopl asmic signaling molecules not drawn to scale. Note that the p85 subunit of PI 3-kinase has been presented in the adaptor category in the absence of the bound pl 10 catalytic subunit. P indicates that the protein is phosphorylated in that region, PTPase is a phosphatase domain and BPS is a putative binding dornain that exists in Grb7 family proteins ktween the EH and SH2 domains. Adapted from Pawson, 1995. Enzymes

A unique class of RTK targets possesses enzymatic functions that may be activated or inhibited by tyrosine phosphorylation or allosteric changes. =domain- containing protein tyrosine phosphatase-2 (Shp2) is a cytoplasmic signaling enzyme that consists of two tandem SH2 domains followed by a potein tyosine ghosphatase (ETP) domain (Figure 1.5). Insight into the mechanism of activation of this enzyme has been gained through determination of the three-dimensional structure of Shp2. Although both

SH2 domains are required for optimal binding to bisphosphorylated targets (Hof et al.,

1998; Ottinger et al., 1998), the amino-terminal SH2 domain functions as a conformational switch to sterically regulate the enzyme by exposing its upon

binding to a phosphoprotein (Hof er al., 1998). These findings demonstrate a mechanism

for catalytic regulation of an enzyme according to its SH2 domain occupancy. A similar allosteric mechanism for regulating enzymatic activity is also proposed for phosphatidyl~nositol3' (PI 3)-kinase. PI 3-kinase is a Iipid kinase that is unique in that it

is cornposed of two separate, constitutively associated subunits: a p 1 IO catalytic subunit

and a p85 regulatory subunit that acts as an adaptor protein, utilizing two tandem SH2 domains to interact with phosphoproteins (Figure 1 S). Tyrosine phosphorylation of p85

has been reported in instances when it associates with tyrosine kinases although there is

no evidence for a functional role for this as PI 3-kinase activation also occurs in the absence of p85 phosphorylation. Instead, binding of specific phosphoproteins to the p85 subunit appears to induce a change in SH2 domain structure and this is transmitted to the

p 1 10 subunit (Shoelson et al., 1993). Alternatively, RTK activation may facilitate subcellular relocalization of an enzyme that places it in close proximity to a physiological target in or near the plasma membrane. For instance, p2lk GTPase-activating protein (RasGAP) is a signaling enzyme that has a carboxy-terminal catalytic subunit that is preceded by an amino- terminal SH3 domain flanked by two SH2 domains (Figure 1.5). RasGAP functions as a negative regulator of Ras signaling by stimulating the intrinsic rate of Ras GTPase activity (Trahey er al., 1988). Upon binding of the SH2 domains of RasGAP to activated

RTKs, RasGAP becomes concentrated at the plasma membrane where it can act on membrane-anchored Ras to catalyze GTP-GDP exchange and thus downregulate Ras

(van der Geer et al., 1997). Taken together. despite the similarities in the overail structure of these cytoplasmic signaling enzymes, it is clear that distinct mechanisms have evolved to control their catalytic activities within the celIular environment.

1.5.3 Intracellular signaling pathways

Cytoplasmic signaling molecules serve to couple activated receptors to downstream biochemical pathways and for the purposes of this discussion. two such pathways will be described here. One of the most established signal transduction pathways is that which initiates Ras activation. As schematically illustrated in Figure 1.6, following growth factor stimulation, the SH2 domain of Grb2 binds to activated growth factor receptors (Lowenstein et al., 1992). However, some RTKs cannot directly bind

Grb2 since they Iack intrinsic Grb2 binding sites and instead, an intermediate molecule such as Shc or Shp2 binds to the receptor and becomes tyrosine phosphorylated, facilitating its interaction with Grb2 (Li er al., 1994; Rozakis-Adcock et al., 1992). Grb2 RTK

Figure 1.6: A schematic representation of the Ras-MAPK and PI 3-kinase signal transduction pathways initiated upon ligand binding and activation of RTKs as descri bed in the text. P represents a phosphorylation event. Adapted from references in the text. interacts constitutively via its SH3 dornains with proline rich regions of the guanine nucleotide exchange factor Son of Sevenless (SOS)(Li et al., 1993; Rozakis-Adcock et al.. 1993). Recruitment of the Grb2-SOScomplex to the plasma membrane localizes SOS near its physiological target Ras which is anchored to the inner face of the membrane (Li er trl.. 1993; Rozakis-Adcock et al., 1993) and SOSactivates Ras by converting Ras from its inactive GDP-bound state to its active GTP-bound state. Activated Ras can then trigger the mitogen xtivated protein kinase (MAPK) pathway which is actually a cascade of three protein kinases that are activated in series (reviewed in Marshall, 1995). Ras first associates with the serine kinase Raf (also known as mitogen wtivated protein

-kinase kinase kinase or MAPKKK) at the plasma membrane. Raf then catalyzes the phosphorylation of mitogen xtivated protein kinase kinase (MAPKK) on serines in its activation loop. MAPKK is a dual specificity kinase that in turn phosphorylates the terminal MAPK on both a tyrosine and a threonine residue in its activation loop.

Activated MAPK is translocated to the nucleus where it can phosphorylate and activate transcription factors such as Elk-l that ultimately effect changes in the pattern of gene transcription such as those involved in the control of ce11 growth and proliferation.

Recent advances in understanding the molecular mechanisms that control programmed ce11 death have identified the PI 3-kinase signal transduction cascade as a critical mediator of extracellular survival signals (reviewed in Downward, 1998). As illustrated in Figure 1 -6, PI 3-kinase becomes activated upon binding to phosphorylated

RTKs via the interactions of the SH2 domains of the p85 subunit. Following the generation of membrane-associated 3' phosphoinositides (3' PI) by activated PI 3-kinase, the serine-threonine protein -nase B (PKB, also known as Akt) is translocated to the plasma membrane via the association of its PH domain with the PI 3-kinase lipid products (Andjelkovic et ai., 1997; Dudek et al., 1997; Franke et al., 1997; Frech et ai.,

1997: Klippel et ai., 1997). Membrane-translocated PKB/Akt becomes phosphory lated and subsequently activated by the ~>hosphoinositide-dependentkinases (PDK) 1 and 2

(Alessi et al., 1996; Alessi et al.. 1997: Stokoe et al., 1997)- Activated PKB/Akt can in turn phosphorylate the pro-apoptotic member of the Bcl-2 farnily known as Bad and this results in its inactivation as a resutt of its dissociation from Bcl-x, and reassociation with

14-3-3 (Datta et al., 1997; del Peso et al., 1997). Released B~1-x~can then converge on the apoptotic machinery of the ce11 to suppress ce11 death.

1.6 Signal Transduction Pathways Controlled by Tek

The identification of a unique family of natural agonists and competitive antagonists for the Tek receptor implies that the functions of this receptor are precisely regulated in vivo. However, since these ligands have only recently been discovered. there is little evidence contributing to our understanding of Tek-mediated signal transduction pathways. At the onset of the present set of studies, the angiopoietins had not yet been isolated and Grb2 and Shp2 were the only reported signaling molecules that could associate with Tek (Huang et al., 1995). Both Grb2 and Shp2 appear to function as positive regulators of signal transduction as they have been linked to activation of Ras and MAPK cell growth signaling pathways (Pawson, 1995). However, it was subsequently demonstrated that the angiopoietins do not behave as endothelial ce11 mitogens (Davis et al., 1996; Witzenbichler et al.. t 998) and the functions of Grb2 and Shpî, in angiopoietin-stimulated Tek signaling pathways remain to be determined. Gene targeting studies have revealed that Tek signaling is required for endothelial ce11 survival as well as angiogenic sprout formation and it therefore seemed likely that additional unknown binding partners for Tek participated in these important signal transduction cascades. The present set of studies was aimed at elucidating the moIecular mechanisms that support the putative roles of Tek signaling in endothelial cell survival and in sprouting angiogenesis. 1.7 Rationale

Signal transduction pathways controlied by Tek within the endothelial ceIl Iineage

have been shown through genetic experiments to be crucial for normal embryonic

cardiovascular development. However, the molecufar mechanisrns underlying these Tek-

mediated signaling pathways are unclear since the angiopoietins have remained elusive

until very recently. The aim of this study was to determine exactly how Tek was eliciting

its signais in endothelial cells through the isolation and extensive characterization of additional downstream targets, substrates and/or docking molecules that bind the activated Tek receptor. The elucidation of the signaling pathways that support the function of Tek in endothelial cells will ultirrtately contribute to out- understanding of normal vascular development in both the ernbryo and the adult. as well as in pathological situations such as tumour-associated angiogenesis. 1.8 Thesis Objectives and Organization

1. To identify downstream signaling targets of the phosphorylated Tek receptor, the

yeast two-hybrid system was used to circumvent the requirement for a ligand to

specifically activate the receptor. Concurrent with this ongoing investigation, the

angiopoietins were isolated and Angl was used to examine the signal transduction

pathways mediated by the putative Tek binding partners. The results of this study

rire discussed in Chapter 2- This chapter was the final study published however it is

presented first in this format since the rernaining chapters focus on one particular Tek

signaling molecule. The reader is referred to Chapter 3 for experimental details.

2. One of the putative Tek binding partners identified using the above approach was a

novel docking molecule that we have narned Dok-R. Initial characterization of this

molecule is detailed in Chapters 3 and 4. Chapter 3 was the first study published and

it outlines the yeast two-hybrid screen approach.

3. To investigate a potential biological rote of Dok-R in RTK-mediated signal

transduction in the absence of the angiopoietins, Dok-R function was examined

downstream of an alternate RTK (EGFR). These findings are presented in Chapter 5. 1.9 Attributions

In Chapter 2, the BIAcore analysis measurements were obtained with the collaborative efforts of Mr. Denis Bouchard at the AMGEN Institute, Toronto, Canada.

The Angl-dependent survivai and migration assays presented in Figure 5, panels C and D were generated with the assistance of Zubin Master, a PhD student in Dr. Dan Dumont's laboratory . The data presented in Chapter 2 has ken published in the following form:

Jones, N., Master, Z,, Jones, i., Bouchard, D., Gunji, Y., Sasaki, H., Daly, R.,

Ali tato, K., and Dumont, D. J. ( 1999). Identification of Tek/Tie2 binding partners.

Binding to a multifunctional docking site mediates cell survival and migration. J. Biol.

Chem. 274,30896-30905.

In Chapter 3, cloning of the full-length Dok-R sequence was performed with the assistance of Martin Bissessar, a former techniça1 assistant in Dr. Dan Dumont's laboratory and the immunofluorescence studies depicted in Figure 2C and the in vitro kinase assay presented in Figure 4A were performed with Dr. Dan Dumont. The data presented in Chapter 3 has ken published in the following form:

Jones, N. and Dumont, D. J. (1998a). The TekITie2 receptor signals through a novel Dok-related docking protein, Dok-R. Oncogene. 17, 1097-1 1 OS.

In Chapter 4, the chromosomal mapping experiments were performed by Genome

Systems, Tncorporated in St. Louis, Missouri. USA. The data presented in Chapter 4 has been published in the following forrn: Jones, N. and Dumont, D. J. (f998b). Assignment of the Dokr gene to mouse chromosome 14D2-D3 by fluorescence in situ hybridization. Genomics. 53,4 13-4 14.

In Chapter 5, Table 1 was generated with the collaborative efforts of Mr. Denis

Bouchard and Figure 3 was generated with the assistance of Dr. Dan Dumont. The data presented in Chapter 4 has ken published in the following fom:

Jones, N. and Dumont, D. J- (1999). Recruitment of Dok-R to the EGF receptor through its PTB domain is required for attenuation of Erk LMAPkinase activation. Crcrr.

Biol. Y, 1057- 1060. CHAPTER 2

IDENTIFICATION OF TEK/TIE2 BINDING PARTNERS: BINDING TO A MULTIFUNCTIONAL MCKING SITE MEDIATES CELL SURVIVAL AND MIGRATION.

A version of this chapter is published in The Journal of Biological Chernistry (N. Jones, 2. Master, J. Jones, D. Bouchard, Y. Gunji, H. Sasaki, R. Daly, K. Alita10 and D. J. Dumont, 1999). ABSTRACT

The Tek/Tie2 receptor tyrosine kinase plays a pivota1 role in vascular and hematopoietic development- In order to study the signal transduction pathways that are mediated by this receptor, we have used the yeast two-hybrid system to identify signaling molecules that associate with the phosphorylated Tek receptor. Using this approach, we demonstrate that five molecules, Grb2, Grb7, Grbl4, Shp2 and the p85 subunit of PI 3- kinase can interact with Tek in a phosphotyrosine-dependent manner through their SH2 domains. Mapping of the binding sites of these moiecules on Tek reveals the presence of a multisubstrate docking site in the carboxyl tail of Tek (YHW). Mutation of this site abrogates binding of Grb2 and Grb7 to Tek in vivo and this site is required for tyrosine phosphorylation of Grb7 and p85 in vivo. Furthemore, stimulation of Tek-expressing cells with Angl results in phosphorylation of both Tek and p85 and in activation of endotheIial ceil migration and survival pathways that are dependent in part on PI 3- kinase. Taken together, these results demonstrate that Ang l-induced signaling from the

Tek receptor is mediated by a multifunctional docking site that is responsible for activation of both ceIl migration and cell survival pathways. INTRODUCTION

RTKs are cell-surface proteins that receive cues from extracellular growth factors to ul timately control biological responses such as ce11 growth, differentiation and survival. Signal transduction pathways mediated by RTKs are initiated upon ligand binding with subsequent receptor-mediated dimerization and autophosphorylation on specific tyrosine residues '(van der Geer et al., 1994). These phosphotyroçine residues serve as high affinity binding sites for numerous intracellular signaling molecules that contain SH2 or PTB domains (Pawson and Scott, 1997). Such specialized protein modules are found in various classes of RTK targets including the p85 subunit of the lipid kinase PI 3-kinase, the protein tyrosine phosphatase Shp2, the docking molecules

Shc and IRS- 1 and the adaptor proteins Grb2, Grb7 and Grb 14. Additional rnodular units such as SH3 and PH domains are also found in Grb2 and the Grb7/Grbl4 family respectively. A regulated cascade of protein-protein interactions allows other signaling molecules to be recruited and assernbled into an organized biochemical network that leads to the activation of important signaling pathways within the cell.

Two subfamilies of RTKs have been identified whose expression is alrnost exclusively restricted to cells of the endothelial 1 ineage (Claesson-WeIsh, 1999;

Mustonen and Alitalo, 1995). The first subfamily contains the high affinity VEGF receptors known as VEGFR- 1 (Flt- 1 ), VEGFR-2 (KDR/Flk- 1 ) and VEGFR-3 (FIt-4). To date. there are four distinct isoforrns of VEGF denoted VEGF-A through VEGF-D that bind and activate different subsets of the VEGFRs (Neufeld et al., 1999). The second subfamily consists of the TIE receptors known as Tie (Tiel) and Tek (Tie2) which are structurally dissimilar from the VEGFRs. While the ligand for Tie has yet to be identified, a family of ligands known as the angiopoietins has recently been shown to bind Tek (Davis et al., 1996; Maisonpierre et al., 1997; Valenzuela et al., 1999).

Interestingly, these ligands appear to have opposing actions as Ang 1 and Ang4 stimulate tyrosine phosphorylation of Tek while Ang2 and Ang3 can inhibit this phosphorylation.

Genetic experiments have demonstrated rhat the signal transduction cascades that are mediated by each of these RTKs are distinct and that each pathway is critical for normal embryonic blood vessel devetopment. VEGFR-1 and VEGFR-2 appear to be required during the initial developmental process of blood vessel formation known as vasculogenesis (Fong et al., 1995; Shalaby et al., 1995) whereby endotheIia1 ceIl precursors differentiate and assemble into a primitive vascular network (Risau, 1997).

This network is subsequently remodeled through the sprouting of new capillaries from pre-existing larger vessels by a process known as angiogenesis (Risau, 1997). VEGFR-3,

Tie and Tek have al1 been shown to be required for this later stage of blood vessel maturation (Dumont et al., 1994: Dumont et al., 1998; Puri et al., 1995; Sato et ai-,

1995).

inactivation of Tek signaling pathways results in early embryonic lethality as a result of angiogenic defects throughout the vascular system including a reduction in blood vessel integrity, decreased vessel sprouting and a loss of endothelial cells (Dumont et al., 1994; Maisonpierre et al., 1997; Sato et al., 1995; Suri et al., 1996). These analyses have revealed that activation of Tek by Angl appears to control at least two different signal transduction cascades that can effect changes in angiogenic sprout

formation and survival of endothelial cells. Consistent with the rote of Tek signaling

pathways in vesse1 sprouting, a modified form of Ang I known as Ang 1 * has recently

been shown to initiate endothelial ceIl migration and sprouting in vitro (Koblizek et al.,

19%: Papapetropoulos et al., 1999; Witzenbichler et ai., 1998). Furthemore, Tek has

been shown to signal through a novel PTB domain-containing docking molecule known

as Dok-R (for downstream tyrosine kinases-slated protein) that can associate with

numerous signaling molecules that are thought to be involved in ceIl migration (Jones and Dumont, 1998a). Collectively these results suggest that Tek signaling through Dok-

R may lead to alterations in the intracellular architecture of endothelial cells which is critical for their directed migration during sprouting angiogenesis.

The identification of a unique family of Tek ligands with opposing functions implies that the signal transduction pathways downstream of Tek are tightly regulated, however, these pathways are currently not well understood. The phosphorylated Tek receptor bas been shown to associate with Grb2 and Shp2 in vitro (Huang et al.. 1995) and more recently, Korpelainen et al. ( 1999) have shown that Tek can activate STAT 1,

STAT3 and STATS. Tek has also been shown to initiate signal transduction pathways downstream of PI 3-kinase and Dok-R (Jones and Dumont, 199th; Kontos et al., 19%).

Here we report that Grb7, Grb14 and the p85 subunit of PI 3-kinase, as well as Grb2 and

Shp2, can associate with Tek. These five molecules can interact specifically with the phosphorylated Tek receptor through their SH2 domains in yeast and in vitro. Using synthetic phosphopeptides, we have mapped the potential binding sites of these molecules on Tek and found that one tyrosine residue, Y"OO , may serve as a multisubstrate doçking site. We provide evidence to suggest that Grb2, Grb7 and p85 can associate with Tek in vivo and that these interactions are mediated through this multidocking site. We also show that Tek can use both Grb7 and p85 as substrates and that tyrosine phosphorylation of these proteins is abrogated when Y''Wis mutated.

Furthemore, we dernonstrate that p85 becomes tyrosine phosphorylated following Ang 1 stimulation of Tek and that a PI 3-kinase signaling pathway is required for Angl-induced endothelial ce11 migration and cell survival.

MATERIALS AND METHODS cDNA library construction and yeast îwo-hybrid screening

Construction of the mouse day 12.5 embryonic heart and lung cDNA library and the yeast two-hybrid screening approach have been extensively described in (Jones and

Dumont, lW8a).

Production of GSTjÙsion proteins

-Glutathione 5-transferase (GST) fusion proteins were prepared from Esclzericlzia coli (E. coli) using standard procedures and the recombinant fusion proteins were purified foIlowing imrnobilization on glutathione-sepharose beads (Pharmacia Biotech). Proteins were eluted from the beads upon treatment with free (10 mM) glutathione for 30 min at

4°C. Purified proteins were analyzed by sodium &decyl sulphate-poly-rylamide gel glectrophoresis (SDS-PAGE) followed by Coomassie Blue staining. The concentrations of the proteins were estimated by cornparison with bovine -mm aibumin (BSA) standards.

Real time binding measurements using BIAcore

Arnino-terrninally biotinylated peptides were synthesized by the AMGEN peptide synthesis group (Boulder. CO) and purity was confirmed by mass spectral and arnino acid composition anaiysis. Relative binding of purifieci GST-SH2 domains to biotinylated phosphopeptides was measured using surface plasmon resonance (BIAcore 2000, Biacore

Inc,, Piscataway, NJ). Peptides were coupled to streptavidin-coated sensor surfaces

(Sensor Chip SA) to a density of 400 response units as per manufacturer's instmctions.

GST fusion proteins were serially diluted in running buffer [IO mM HEPES (pH 7.4),

150 mM NaCl, 3 rnM EDTA, 0.005% (v/v) surfactant P20] and injected at a 5 pumin flow rate. Surface regeneration was performed by 10 pL injections of 1 M NaCl in 50 mM NaOH at a 20 pL/min flow rate. Relative binding was measured as an increase in arbi trary response units and kinetic parameters were detennined using the BIA evaiuation

3.0 software.

GST binding and coimmunoprecipitation assays

The cDNAs representing the cytoplasmic domain of Tek, both wildtype and kinase inactive, were subcloned into pACTag2 (Jones and Dumont, 1998a) to generate

-hemagglutinin (HA)-tagged proteins. The cDNA representing the full-length Tek receptor was cloned into the pRc/RSV vector (Invitrogen) and the corresponding full- length ~ek'''~'and ~ek~''* cDN As were cloned into pcDNA3.1(+) (Invitrogen). Grb7 cDNA has been previously described (Fiddes et al., 1998) and p85 cDNA was a gift of

Anke Klippel (Chiron, Emeryville, California). 10 pg of each DNA was used to transfect

a 10 cm culture dish of human embryonic kidney (HEK) 293T cells, using lipofectin

reagent (Gibco BRL), according to the manufacturer's instructions. For Ang 1

stimulations on endothelial cells, cells were serum-starved for 6 h, pre-treated with 1 rnM sodium orthovanadate pH 8.0 for 10 min at 37°C and stimulated with 5 rnL of conditioned medium in the presence of sodium orthovanadate for 10 min. Alternatively.

for hg1 stimulations on HEK293Tek cells, cells were serum-starved for 24 h and were

stimulated in the absence of phosphatase inhibitors. GST mixes, immunoprecipitations and Western blotting were performed as previously described (Jones and Dumont,

1998a).

Peptide association assays

1 pg of eluted fusion protein was incubated with 1 pg of biotinylated peptide in 1%

Triton X-100 lysis buffer for 2 h at 4°C. Complexes were recovered on streptavidin agarose beads (Pierce), eluted in sample buffer, and processed as described in (Jones and

Dumont, 1998a).

A ntibodies used for immunoprecipitation and western blomhg

Commerciatly available antibodies used were as follows: for irnmunoprecipitation, polyclonal anti-Tek C-20 (Santa Cruz), polyclonal anti-Grb7 N-20 (Santa Cruz), monoclonal anti-Myc (Invitrogen) and for Western bIotting, monoclonal anti- phosphotyrosine 4G 10 (Upstate Biotechnology, Inc.), monoclonal anti-HA-horse~adish peroxidase (HRP) clone 12CA5 (Boehringer Mannheim), monoclonal anti-Grb2

(Transduction Laboratories), monoclonal anti-Grb7 (Transduction Laboratories), monoclonal anti-,Mye (Invitrogen), HRP-conjugated donkey anti-human IgG H+L

(Jackson ImmunoResearch). Monoclonal and polyclonal anti-Tek antibodies specific to the extracellular domain were a kind gift of Fu-Kuen Lin (AMGEN, Thousand Oaks,

California). A peptide specific to the kinase insert region of Tek (H,N-

CRKSRVLETDPAFAVANSTAST-COOH) was used to raise a potyclonal anti-Tek antisemm in rabbits and this affinity purified antibody is referred to as anti-TekK'.

Potyclonal anti-p8S antibody was generously provided by Anke Klippel.

Ce11 culîure and production of condiîïoned medium

Passage 15 human umbilical -in gndothelial -11s (HUVEC) were obtained from ATCC and cuitured in F12 medium supplemented with 15% etal bovine -mm (FBS), 1% penicillin. 1% streptomycin, 200 mM L-glutamine (al1 Gibco BRL), 0.I mg/rnL heparin

(ICN Biomedicals) and 0.02 mg1mL bovine endothelid ce11 growth factor (ECGF)

(Boehringer Mannheim) in a 37"C, 5% CO2 incubator. HEK293. HEK293T,

HEK293Tek (a gift of Fu-Kuen Lin), Py4- 1 (a gift of V. Bautch, North Carolina) and

EA.hy926 (a gift of Cora Edgell, North Carolina) were grown on 10 cm plates in

-Dulbecco's modified Eagle's medium (DMEM) (Gi bco BRL) supplemented with 10%

FBS, 1 % penicillin, 1% streptomycin, and 200 mM L-glutamine. HEK293Tek cells were further supplemented with 250 pg/mL G418 (Gibco BRL) and 100 nM methotrexate

(Sigma) and EA.hy926 were further supplemented with HAT (hypoxanthine, aminopterin and thymidine) (Sigma). Ang l cDNA was subcioned into either SignalpIg-plus (Novagen) or pSecTagB (Invitrogen) and HEK293T cells stably expressing Ang 1-F, or

Ang 1 -MH were generated by selection of transfected cells with either 1.5 mg/mL G4 1 8

(Gibco BRL) or 1 mg/rnL zeocin (Invitrogen) respectively and resistant cells were

pooled. Non-transfected HEK293T cells and stable transfectants were grown to confluence on 15 cm dishes and conditioned medium was collected for 24 h in DMEM

supplemented with O. 1 % or IO% FBS. Harvested medium was clarified by centrifugation and secretion of Ang 1-F, or AngI -MH into the medium was confinned by Western analysis. Ang 1-F, was depleted from 10 rnL of conditioned medium by incubation with

500 pL of 20% protein A-sepharose slurry for 1 h at 4°C.

Cell migration assay

HUVEC, Py4-1, HEK293 and HEK293Tek cells were seeded at a density of 8.4 x 104 cells in 500 pL of DMEM + O. 1 % FBS, with or without inhibitor, in the upper chamber of an 8 pm-pore modified Boyden charnber (Falcon). Conditioned medium was placed in the lower charnber and wortmannin or LY299402 (both Sigma) were added at a final concentration of either IO, 50 or 100 nM or 10, 20 or 40 pM respectively. Cells were allowed to migrate for 4 h in a 37"C, 5% CO, incubator. Non-migrating cells were scraped off and filters were fixed in 100% methanol for 5 min, stained with Harris'

Hematoxylin (BDH) for 10 min and washed twice with tap water for 3 min each. Filters

were then mounted using 100% glycerol (Fisher Biotech) and counted using a

microscope (Leica) at 4ûûx magnification. Endothelia1 ce11 survival assay

HUVEC were seeded at 1.2 x 10' cells/well in 24-well dishes in DMEM +10% FBS.

After 24 h in culture, cells were washed once in lx phosphate buffered sline (PBS) and

incubated in 1 mL of fresh DMEM + IO% FBS or Mock, Ang 1-Fc,Depleted or Ang t -Fc

+ 10 nM wortmannin conditioned medium. After 5 days of incubation, viable cells were

identified by Trypan-blue exclusion (Gibco BRL) and counted using a hemacytometer

(Bright-Line).

Statistical analysis and image processing

Statistical significance was detennined using Student's t test for comparisons between

two means. Al1 experiments were performed in triplicate. A p value of Iess than 0.05 a was interpreted as statistically significant. Results are graphically expressed as the mean

+/- standard error of the mean. Gels were digitally scanned and processed using

Photoshop 5.0 on a Macintosh G3 cornputer.

RESULTS

Identification of Tek binding partners

To identify proteins that could participate in Tek signaling pathways, we have

used the yeast two-hybrid system to fînd molecules that associate with the cytoplasmic

domain of Tek in a phosphotyrosine-dependent manner (Chien et al., 199 1 ; Fields and

Song, 1989). Using this approach, we have previously reponed the identification of Dok-

R as a Tek binding partner (Jones and 'Dumont, 1998a). Initially, we tested whether Tek

could associate with two known binding partners, Grb2 and Shp2 (Huang et al., 1995), in a phosphotyrosine-dependent manner in yeast. The entire intracellular domain of either the wildtype Tek receptor (~ek"), the catalytically inactive Tek receptor (~ek~~~~'~)or a mutant receptor bearing a tyrosine to phenylalanine mutation in the predicted Grb2 in the tail (TekFitm') (Huang et al.. 1995) were coexpressed in yeast with the

SH2 domains of Grb2 and Shp2. Expression of the truncated receptors in yeast results in constitutive tyrosine phosphorylation of both Tekrc and TekF'lm" while TekAgs3'"remains unphosphorylated (Figure 2.IA). Yeast expressing both Tek1' and Grb2 were able to activate the lad reporter gene and grow on selective medium (Figure 2.1 B and data not shown). In contrat. yeast expressing either TekAgS31Cand Grb2 or TekF"*iC and Grb2 produced very low B-galactosidase activity and sparse growth on selective medium. illustrating that this interaction was dependent upon Tek kinase activity and an intact

Y''*residue (Figure 2.1 B and data not shown). Sirnilar results were obtained with Shp2

(data not shown), indicating that the yeast two-hybrid system could be used to identify putative Tek binding partners.

A yeast strain expressing Tekrc was then used to screen for interacting proteins expressed from cDNAs obtained from a day 12.5 mouse embryonic heart and lung library as described previously (Jones and Dumont, 1998a). Early embryonic heart and lung express high IeveIs of Tek (Dumont et al.. 1992) thus it was reasoned that cDNAs derived from these tissues should be rich in Tek signaling partners. Several clones were identified that interacted with Tek in a phosphotyrosine-dependent manner. Figure 2.1 C lists the names of the signaling molecules that were either obtained in the screen or were shown to interact with Tekrc or Tekic mutants in yeast. Grb2, Shp2 and p85 have IP: a-Tek

Figure 2.1: Expression of Tek in yeast and cloning of Tek binding partners. (A) Whole ce11 yeast lysates were analyzed for Tek expression and tyrosine phosphorylation usine an ti-Tek and an ti-phusphotyrosine (pY) an tibodies. Lysates were prepared from yeast expressing the intracellular domains of either wildtype Tek (TektC), kinase inactive Tek (TekA8531C ) or Tek beanng a tyrosine to phenylalanine mutation in the predicted Grb2 binding site in the carboxyl tail (Tek F""'C ). AI1 intracellular domains are expressed and ~ek"and are tyrosine phosphorylated. (B) Yeast coexpressing the SH2 domain of Grb2 and TekIc, Tek A853'C or ~ek~''m'C were assayed for lac2 reporter gene activation. Only yeast expressing both Grb2 and wildtype T'ePC were able to activate the reporter gene as indicated by a chromogenic assay (blue precipitate in yeast). (C) Summary of Tek binding partners isolated in yeast two-hybrid screen or found to associate with Tek in yeast. For Shp2, the individual SH2 domains (N or C) or the tandem SH2 domains (N+C) were introduced into yeast separately. The relative intensities of the interactions between the putative Tek binding partners and the various Tek intracellular domains are indicated with + or -. previously ken reported to serve as targets for numerous activated RTKs while Grb7 and

Grb 14 are members of an emerging farnily of PH domain-containing adaptor rnolecules.

Importantly, we confirmed that al1 of these signaling rnolecules are coexpressed with Tek

in cultured endothelid cells (data not shown), suggesting that Grb2, Grb7, Grb 14, Shp2 and p85 could indeed be true signaling partners of Tek.

The SH2 domains of Grb2, Grb7, GrbZ4, Shp2 andp85 mediate biding ta Tek

Sequence analysis of the partial Grb7, Grbl4 and p85 cDNAs obtained in the screen demonstrated that they contained the regions coding for the SH2 domains of these proteins. To determine whether it was indeed the SH2 domains that mediated these interactions, and also to test whether these interactions could occur outside of the yeast environment, we fused the SH2 domains of Grb2, Grb7, Grbl4, Shp2 and p85 to GST and purified these fusion proteins from E. di. Immobilized GST-SH2 domains were incubated with lysates prepared from HEK293T cells transiently expressing the HA- tagged cytoplasmic domains of either wildtype Tek (Tek1=)or catalytically inactive Tek

(TekhRsMc).In a manner similar to that seen in yeast, overexpression of Tek1=in mammalian cells results in its autophosphorylation and the kinase inactive form of ~ek" remains unphosphorylated (Figure 2.2A). The isolated SH2 domains of Grb2, Grb7 and

Grb14 could al1 precipitate Tek", but were unable to precipitate TekA"'" (Figure 2-28).

Both Shp2 and p85 contain two SH2 domains thus we constructed GST-SH2 fusions of each domain individuaily to determine which of these domains mediated the interaction with Tek. Both SH2 domains of Shp2 and p85 were able to interact with Tektc; however, the carboxy (C)-terminal SH2 domain of Shp2 was the least effective at precipitating Blot: a-Tek a-pY

Figure 2.2: In vitro association between Tek and the various SH2 domains of putative Tek binding paruiers. (A) Lysates from 293T cells expressing a tagged, uuncated fom of Tek (~ek")or kinase inactive Tek (~ek*"~'~)were immunoprecipitated with anti-Tek C20 antibodies. Expression of ~ek'~in 2931 cells results in its tyrosine phosphorylation whereas no tyrosine phosphorylation of ~ep~~~~~was detected. The migration of ~ek" and ~ek~~~'~~is indicated with an arrow. (B) Lysates from (A) were incubated with immobilized GST-SH2 domains from the various Tek binding partners or GST done. Resulting complexes were resolved by SDS-PAGE and immunoblotted with anti- hemagglutinin-horseradish (HA-HRP) peroxidase. Strong associations were detected with the SH2 domains of Grb2, Grb7, Grbl4 and p85 whereas no associations were detected with GST alone. The amino-terminal SH2 domain of Shp2 exhibited strong binding similar to the tandem N+C SH2 domains while the carboxy-terminal SH2 domain exhibited weak binding. Tek1' (Figure 2.2B). These in vitro binding data supported our results obtained in yeast

(Figure 2. IC) where the Shp2 arnino (N)-terminal SH2 domain mediated stronger binding

to Tekic than the carboxy-terminal SH2 domain. Taken together. these results

demonstrate that the SH2 domains of Grb2, Grb7, Grb 14 and p85 and the amino-terminal

SH2 domain of Shp2 can al1 bind specifically to Tek in vitro and that these interactions

are dependent on Tek tyrosine kinase activity.

Defermination of the receptor binding sites in vitro

To this point, we have shown that the SH2 domains of these signaling molecules can interact with the receptor in a phosphotyrosine-dependent manner both in yeast and

in lvirru and we next wanted to detennine the putative binding sites of these SH2 domains on Tek. Since the in vivo autophosphoryiation sites of Tek have yet to be defined, we

dssigned synthetic phosphopeptides representing al1 19 possible tyrosine phosphorylation sites on Tek (Figure 2.3A). Aithough it is highly unlikely that al1 of these tyrosine

res idues are phosphory lated in response to Ang 1 stimulation, these phosphopeptides were

used as a starting point to reveal the sites that could potentially mediate an interaction.

The interactions between the various GST-SH2 domains and the 19 potential binding

sites were investigated using real time biosensor (BIAcore) analysis where the

phosphopeptides were immobilized on sensor chips. GST alone did not bind strongly to

any of the immobilized phosphopeptides while the SH2 domains of both Grb2 and Grb7

bound most strongly to the peptide comprising phosphorylated tyrosine residue (pY)

1 100 (pY1lW)and not to the unphosphorylated counterpart (Figure 2.3B and data not

shown). pY1Imis within a Y"% context and this sequence fits the consensus binding pY814-NPDPTlpYPVLDW pY858-IKRMKEpYASKDDH pY895-ACEHRGpYLYLAIE pY897-EHRGYLpYLAlEYA pY902-LYLAIEpYAPHGNL pY952-VARGMDpYLSQKQF pY974-I LVGENpYIAKIAD pY990SRGQEVpWKKTMG pY 1O1 O-AlESLNpYSVYTTN pY 1O1 3-SLNYSVpYTTNNSD pY 1022-NSDVWSpYGVLLM pY 1037SLGGTPpYCGMTCA pY 1046-MTCAELpYEKLPQG pY 1053-EKLPQGpYRLEKPL pY 1066-NCDDNpYDLMRQC pY 1078-CWREKPpYERPSFA pY1100-MLEERKTpWNTTLYE pY 1 106-VNTTLpYEKFTY pY1111-EKFTpYAGIDCS

Figure 2.3: Mapping of the SH2 domain binding sites on Tek using phosphopeptide analyses. (A) Schematic representation of the intracellular domain of the mouse Tek receptor showing the 19 putative tyrosine phosphorylation sites and the corresponding synthetic phosphopeptides that were used in the mapping studies. TM indicates the transmembrane domain while TKI and TK2 represent the two lobes of the kinase domain. Note the absence of tyrosine residues in the region that separates TKl from - TK2. Amino acid residues are designated according to single-letter code and the position of the phosphotyrosine (pY) is in bold type. (B) Graphic representation of the results obtained from BIAcore analysis using purified GST fusion proteins and immobilized phosphopeptides. Relative response units were measured and interpreted from sensograms. (C) Associations between various Tek phosphopeptides and purified GST- fusion proteins were assayed in vitro. The interactions tested are summarized and relative binding is indicated with + or -. OOOOOO OOO~O OOOOO VIwON- site for the SH2 domains of both Grb2 and Grb7 that predicts the presence of asparagine at the 1-2 position following the tyrosine (Janes et al., 1997; Songyang et al., 1993;

Yokote et al., 1996). In contrast to Grb2 and Grb7, the SH2 domain of Grb 14 did not bind strongly to pY1lW,rather this SH2 domain was found to bind most effectively to peptides representing pY"" and pY'lo6 (Figure 2-38). pY8" is found within the context

YXi4PVLwhereas PY"~is found within the context Y'IMEKF, suggesting that Grb14 may bind preferentially to sites where the +3 position is occupied by a large hydrophobic residue.

The binding of Shp2 to these immobilized peptides was somewhat more complex since peptides representing pYa4, pYm5. pY8", pYIO1'. pYI0" and pY"" were al1 found to interact strongly with the Shp2 tandem SH2 domains (Figure 2.3B). These residues are found within the contexts Y8'"PVL, Y"'LYL, Y~~~LAI,YI0I0SVY. Y'02'GVL and

Y""AGI respectively. Interestingly. when the individual SH2 domains were tested in these binding expex-iments, the arnino-terminal SH2 domain was able to bind to nearly al1 of the tyrosine residues initially detected using the tandem SH3 domains while the carboxy-terminal SH2 domain bound very poorly to these residues (Figure 2.38). This result suggested that the binding detected when using the tandem SH2 domains can be mostly attributed to the amino-terminal SH2 domain. The consensus binding site for the

Shp2 amino-terminal SH2 domain predicts a hydrophobic valine or isoleucine at the +1 and +3 positions following the tyrosine (Songyang et al., 1993) and these tyrosine residues conform closely to this consensus. In contrast to Shp2, when the individual SH2 domains of p85 were tested, the carboxy-terminal SH2 domain displayed consistently higher binding to the vanous phosphopeptides than the amino-terminal SH2 domain

(Figure 2.3B). The p85 SH2 domains were found to bind to peptides representinp p~895, p~'037.p~"~~, pY''Mand pYiioaand these residues are within the contexts Y8"~YL,

Y'O"CGM. YlW~LM,Y"9NT and Y1'&EKFrespectively. The consensus binding site for both the amino- and carboxy-terminal SH2 domains of p85 predicts the presence of methionine at the +3 position following the tyrosine (Songyang et al., 1993) and both

Y and Y'- are followed by this predicted consensus.

To determine whether the interactions detected by BIAcore analysis were sufficient to allow coprecipitation, we performed in vitro mixing experiments with purified GST-SH2 domain fusion proteins and the various phosphopeptides. BIAcore analysis indicated that the SH2 domains of both Grb2 and Grb7 bound most strongly to the peptide representing p~''~and these findings were confirmed in the in vitro mixing experiments (Figure 2.3C). Collectively. these results demonstrate that Grb2 and Grb7 can associate with the same tyrosine residue located in the carboxy-terminal tail of Tek.

In contrast. Grbl4 appeared to bind preferentialiy to pY8" in the juxtamembrane region of the receptor (Figure 2.3C), supporting the data from the BIAcore analysis that the

Grb 14 SH2 domain requires the presence of a hydrophobic residue at the +3 position.

In the case of Shp2, the tandem SH2 domains exhibited a complex binding pattern upon BIAcore analysis and similar results were obtained in the in vitro coprecipitation experiments (Figure 2.3C). The tandem and amino-terminal SH2 domains could be coprecipitated with the p~814,pygg5, p~~~~, p~'022and p~""peptides whereas the carboxy-terminal SH2 domain could only be coprecipitated by the pYiO" and pYi'"

peptides. However, Y"', and Yion are found within the conserved activation and

catalytic loops of the Tek kinase dornain, suggesting that they probabiy do not confer

receptor binding specificity in vivo. Taken together, the results from the BIAcore

analysis and the in vitro coprecipitation experiments demonstrate that the majority of

binding of Shp2 to Tek is mediated by the arnino-terminal SH2 dornain and this

interaction is likely through a consensus Shp2 binding motif in either the juxtamernbrane or carboxy-terminal tail of the receptor.

Lastly, coprecipitation experiments demonstrated that the amino-terminal SH2 domain of p85 could interact with pYiM6and pYiiiipeptides while the carboxy-terminal

SH2 domain could associate with pYiO", pYiloOand pY"06 peptides (Figure 2.3C).

Although YtO" and yimsare followed by the predicted p85 SH2 binding consensus, these residues are also found within the conserved kinase domain of the receptor. Colfectively, these experiments demonstrate that the interaction of p85 with Tek is likely mediated through the carboxy-terminal SH2 domain of p85 and a non-consensus binding motif in the carboxy-terminal tail of Tek.

These analyses have demonstrated that the Grb2, Grb7 and p85C SH2 domains al1 associate mosc strongly with the PY"~peptide (Figures 2.3B and C). suggesting that these three signaling molecules may compete for the same binding site on Tek. Using

BIAcore software, we have determined the relative dissociation rate constants (K,) for each SH2 domain binding to the pY'lmpeptide. The SH2 domains of Grb2, Grb7 and p85 were found to bind to p~"~with apparent rate constants of II, 9 1 and 38 nM respectively (Table 2.1). These values demonstrate chat the SH2 domain of Grb2 has the greatest relative affinity for whereas that of Grb7 has the lowest relative affinity for this phosphopeptide.

TABLE 2.1

Apparent equilibrium dissociation constants of GST-SH2 fusion proteins binding to immobilized Tek pY"m phosphopeptide using BIAcore analysis.

GST-SH2 fusion protein Binding constant (K,) In vivo assoc~tionof Grb2, Gr67 and p8S with Tek in 2937 cells

To study the signal transduction pathways mediated by the Tek receptor in the absence of Angl, we first engineered rnammalian expression vectors to contain full- length Tek, TekAs5' or ~ek~'"".When these proteins are transiently expressed in

HEK293T cells, Tek becomes highly tyrosine phosphorylated while TekAM3remains unphosphorylated (Figure 2.4A). TekF1IWis only weakly tyrosine phosphorylated compared to Tek (Figure 2.4Aj; however, we have been able to show that mutagenesis of this site does not affect the catalytic function of the receptor since Dok-R can still become tyrosine phosphorylated when coexpressed with this mutant (Jones and Dumont, unpublished data). These results suggest that Y"OO may represent a major autophosphorylation site on Tek in vivo although there are additional sites of tyrosine autophosphorylation.

To determine whether the associations we had detected in yeast and in vitro could occur irr vivo, coimmunoprecipitation experirnents were performed using lysates prepared from these transfected cells. Coimmunoprecipitation of endogenous Grb2 with Tek is only observed when Tek is tyrosine phosphorylated (Figure 2.4B), demonstrating that

Grb2 cm associate with Tek in vivo and that this interaction is dependent on Tek kinase activity. Furthermore, this interaction does not occur when Y"" is mutated (Figure

3.4B), illustrating that this site is required for the association of Tek with Grb2.

Similarly, Grb7 is coexpressed with either Tek or Tek mutants and coimmunoprecipitation of Tek with Grb7 is also only observed when Tek is phosphorylated at Y'"'(' (Figure 2.4C and data not shown). Furthermore, Grb7 becomes IP: a-Tek a-Grb2- 9 - IP: a-Tek - Lysate

C D

Blot: Biot: 194- a-pY 0 +p8S +Tek 120- : t IP: a-p85 a-pY a-p85

87- * r 4-pp85 Lysate +PP~O 64- +Grb7 11) 00- IP: a-Grb7

a-Grb7 ----Lysate

Figure 2.4: Zn vivo associations between Tek and putative downstream binding partners. (A) Full-length Tek, ~ek*") or ~ek~""were transientiy expressed in 293T cells and lysates were immunoprecipitated with ~ek". (B) Tek immunoprecipitates from above were resolved and irnmunoblotted with anti-Grb2 antibodies. (C) Tek, el?'^ and -rekF 1 100 were coexpressed with Grb7 and lysates from these cells were immunoprecipitated with anti-Grb7 N-20 antibodies. (D)p85 was coexpressed with Tek or Tek mutants and lysates were immunoprecipitated with anti-p85 antibodies and immunoblotted with anti-phosphotyrosine antibodies. Equal expression of Grb2, Grb7 and p85 was detected in al1 lysates. detectably tyrosine phosphorylated when coexpressed with wildtype Tek and this results

in the association of Grb7 with two other phosphoproteins with relative sizes of

approximately 70 kDa (pp70) and 85 kDa (pp85) (Figure 2.4C). It has been reported that

Grb7 can associate with Shp2 which has a relative molecular mass of 70 kDa (Keegan

and Cooper, 1996) and phosphorylated Shp2 may also associate with the p85 subunit of

PI 3-kinase (Welham et al., 1994). However, these two proteins did not react with anti-

Shp2 or anti-p85 antibodies (data not shown) and it remains to be detennined whether these proteins are either Shp2 or p85 or novel Grb7 binding proteins. Upon coexpression of the p85 subunit of PI 3-kinase with Tek or Tek mutants, immunoprecipitated p85 is tyrosine phosphorylated and this phosphorylation is also dependent on Tek kinase activity and an intact Y'IWsite (Figure 2.4D). Because we could not detect Tek in p85

immunoprecipitates (data not shown), it is possible that the interaction between these two proteins may be either transient or indirect. In similar studies, we were unable to detect any in vivo associations of Tek with either Grb14 or Shp2 (data not shown). In summary. these results demonstrate that a multisubstrate docking site on Tek. Y 'lm. mediates an association with Grb2 and Gr67 in vivo and that phosphorylation at this site is required

for tyrosine phosphorylation of Grb7 and p85.

A ngl induces PI 3-kinase-dependent endothelia1 ce11 migration in vivo

While these experiments were ongoing, Angl was identified as an activating

ligand for Tek (Davis et al., 1996). To obtain sufficient quantities of this ligand for

biochemical manipulation, we generated two HEK293T ceIl lines stably secreting Ang 1

and to facilitate the identification of the ligand in these cell lines, Ang 1 was tagged at the carboxy-terminus with either a Myc epitope and poly-histidine (Angl-MH) or the F, region of human irnmunoglobulin (Ang 1 -FJ. Since this ligand is reportedly difficult to purify in its native state (Davis et al., 1996), we proceeded to use Angl-containing conditioned medium collected from these cells in our experiments. Similar expression levels of Ang 1-MH and Ang 1 -F, can be detected in conditioned medium harvested from transfected cells but not from non-transfected control cells (Mock) and Angl-F, can be specifically depleted from the conditioned medium using protein A-sepharose (Depleted)

(Figure MA). Moreover, we have previously shown that both Angl ligands function in a similar manner in in vitro studies (Teichert-Kuliszewska et al., 2000).

We tested the bioactivity of the native ligand on various endothelial ce11 lines that have been shown to express Tek and found that, in the presence of sodium orthovanadate, conditioned medium containing Ang 1 -MH could stimulate tyrosine phosphorylation of

Tek (Figure 2.5B). However, when we Iooked at various markers of cellular activation, we found that sodium orthovanadate alone had many of the same effects as Angl-MH conditioned medium (data not shown). To effectively study the intracellular changes caused by AngI in the absence of sodium orthovanadate, we have used an HEK293 ce11 line that stably expresses the full-length Tek receptor (HEiC293Tek). Interestingly, stable overexpression of Tek in these cells does not result in constitutive activation of the receptor and we are able to induce tyrosine phosphorylation of Tek with Angl-MH conditioned medium without any addition of sodium orthovanadate (Figure 2.5B).

Furthemore, we found that p85 transiently expressed in HEK293Tek cells could become tyrosine phosphorylated following Angl stimulation (Figure 2.5B). Collectively these 97 - 66- me= IP: u-Myc Blot: a-Myc

Blot: a-pY u-Tek

Figure 2.5: Ang 1 stimulates PI 3-kinase-dependent endothelial ce11 migration and survival. (A) Ce11 pellets and conditioned medium were immunoprecipitated with anti- Myc antibodies or precipitated with protein A-sepharose (Pr. A) followed by immunoblotting with either anti-Myc or donkey anti-human IgG (a-F,)respectively. (B) Lysates rrom Py4-1, a polyoma transformeci munne endothelial cell line and EA.hy926, an imrnortalized HUVEC line, were stirnulated as indicated in the presence of orthovanadate or lysates from HEK293Tek cells stimulated in the absence of orthovanadate were immunoprecipi tated with an ti-Tek polyclonal an tibodies and immunoblotted with anti-phosphotyrosine or anti-Tek antibodies. p85 was uansiently expressed in HEK293Tek cells and immunoprecipitaied with anti-p85 antibodies followed by immunoblotting with anti-phosphotyrosine antibodies. 40

20 orVEGF 'FGF :i -FC Wort. Wort. WOR. LY LY LY Growth Condition

Growth Condition

Figure 2.5: (C) The chernotactic response of HUVEC. Py4- 1, HEK293Tek and HEK293 was assessed using a Boyden chamber apparatus. The PI 3-kinase inhibitors wortrnannin (Wort) and LY294002 (LY) have been added where indicated. Data points for HUVEC and Py4-1 represent the average nurnber of cells migrated relative to VEGF treatment whereas HEK293Tek and HEK293 were compared to HEK293Tek treated with Angl-F,. Al1 experiments were performed in triplicate and differences were found io be statistically significant @<0.05). (D) HUVEC were cultured in DMEM with 10% semm supplemented with ECGF, Mock, Angl-F,, Depleted or Angl-F, + 10 nM Wort conditioned medium. Statistical analysis revealed that al1 differences were statistically significant- results show that the intact Angl ligand can induce tyrosine phosphorylation of p85 downstrearn of the Tek receptor.

Ce11 migration has previously been shown to require PI 3-kinase signal transduction pathways (Kundra et al., 1994) thus we set out to determine whether Angl could mediate endothelial ceIl migration through PI 3-kinase. A modified form of AngI

(Ang 1") has previously been reported to be a chemotactic agent for endothelial cells

(Witzenbichler et al., 1998). Using a modified Boyden chamber approach, we assessed the chemotactic response of HUVEC, Py4-1, HEK293Tek and HEK293 cells to conditioned medium containing Ang 1-F, collected in low serurn. Ang 1-F, medium produced a significant migration of HUVEC, Py4-1 and HEK293Tek cells when compared to Mock or Depleted medium and HEK293 cells displayed a weaker response to Ang 1-F, than HEKS93Tek (Figure 2.5C). Interestingly, the PI 3-kinase inhibitors wortmannin and LY294002 could significantly block, although not completely, the chemotactic responses of Angl -F, (Figure 2.5C). These results clearly demonstrate the ability of Ang 1 to mediate chemotaxis of Tek-expressing endothelial and non-endothelial cells and they support a role for PI 3-kinase activity in endothelial cell migration.

Angl funciions as a survival factor for endothelial cells

Signal transduction pathways that regulate the process of cell survival are beginning to be elucidated and PI 3-kinase has been shown to promote ceIl survival through activation of the PKB/Akt pathway (Downward, 1998). To assess whether Angl could initiate PI 3-kinase-dependent survival pathways downstream of Tek, HUVEC were seeded in conditioned medium containing Angl-F, collected in 10% serum and following 5 days in culture, the number of viable cells were counted. Endothelid cells cultured in mock medium exhibited a rnarked decrease in the number of cells surviving after 5 days; however, Angl-F, was able to prevent cell death in HUVEC albeit at lower levels than those seen in cells cultured in medium containing ECGF (Figure 2.5D).

De plet ion of the Ang 1-F, from the conditioned medium abolished this Ang 1-F,- dependent ceIl survival to levels that were equivaient to cells grown in mock medium

(Figure 2.5D). The addition of wortmannin to the Angl-F, conditioned medium also resulted in a dramatic reduction in cell survival (Figure 2.5D), demonstrating that activation of the PI 3-kinase pathway may also be required for Angl-dependent endothelial cell survival.

DISCUSSION

In this study, we report the identification of a nurnber of SH2 domain-containing signaling molecules, namely Grb2, Grb7, Grbl4, Shp2 and p85, that can associate specifically with phosphorylated Tek in yeast and in vitro. We have rnapped potential binding sites of the SH2 domains of these molecuies on Tek using synthetic phosphopeptides and found that one tyrosine residue, YtIm,can serve as a multifunctional docking site. This multisubstrate docking site is required for the association of Tek with

Grb2 and Grb7 in vivo and tyrosine phosphorylation of Grb7 and p85 is abrogated when yi ioo is mutated. Furthermore, tyrosine phosphorylation of both Tek and p85 is rapidly derected following Ang 1 stimulation and PI 3-kinase activity is required in part for Ang1- induced endothelial cell migration and survival. In summary, these results present evidence to suggest that signaling from Tek is mediated by a host of signaiing molecules that can associate with Tek through a multifunctional docking site that is ultimately responsible for activation of endothelid ce11 migration and survival pathways.

The Tek binding partners that were identified in this screen contain SH2 domains that we have shown to be essential for the interaction of these proteins with phosphorylated Tek in virro. Moreover, we have identified putative key phosphotyrosine residues on Tek that mediate binding to the SH2 domains of these proteins. Gr62 and

Grb7 bind almost exclusively to Y"" located in the carboxy-terminal tail of Tek while

Grb 14 interacts preferentially with Y'" and YHo6in the juxtamembrane and carboxy- terminal tail regions respectively (Figure 2.6). Although the binding specificity of the

Grb 14 SH2 domain has yet to be determined, these findings suggest that the preferred binding site for the SH2 domain of Grbl4 may include the presence of a large hydrophobie residue at the +3 position.

Binding studies on proteins such as Shp2 and p85 that contain tandem SH2 domüins have suggested that between the two domains confers high biological specificity (Ottinger et al., 1998). Borh SH2 domains of p85 are required for a stable interaction (Cooper and Kashishian, 1993) although the carboxy-terminal SH2 domain has been shown to be responsible for the high affinity and high specificity interaction of p85 with the PDGF receptor (Klippel et al., 1992). The distance between p85 binding motifs in bisphosphorylated target peptides can be quite short (Ottinger et cd., 1998) and based on Our binding data, it is possible that the carboxy-terminal SH2 1111 ( Endothelial Cell Migration Endothelial Cell Suwival

Figure 2.6: Summary of putative docking sites on Tek for SH2 domain-containing Tek binding panners. Grb14 and ShpZ both interact with Ys'' in the juxtamembrane region and YHo6and Y"ll respectively in the carboxy-terminal tail region of Tek. A multifunctional docking site. YLiW,is also present in the carboxy-terminai tail of Tek and this site is required for binding of Grb2 and Grb7 as well as phosphorylation of both Grb7 and p85. Phosphorylated Grb7 associates with two unknown proteins, pp70 and pp85, and the function of Grb7 remains to be elucidated. In contrast, PI 3-kinase activity, which is likely conferred by p85 binding to is required for Ang 1-induced endothelial ce11 migration and survival. domain of p85 would bind to while the amino-terminal SH2 domain would stabilize

this interaction through binding YtLo60rY"". These data are in agreement with the data

from Kontos et al. (1998) which suggested that p85 binds to Tek through a non- consensus binding motif similar to those seen in other RTKs (Cunningham et al., 1995).

Grb2, Grb7 and p85 were ail found to bind to the same phosphorylated tyrosine residue on Tek (Y"O0) both in vitro and in vivo, suggesting the existence of a multidocking site on Tek. Several precedents have been set in other RTKs where different SH2 domain containing proteins interact with the same phosphotyrosine residue

(Braunger et al., 1997; Ponzetto et al., 1994). The ability of different SH2 domain- containing proteins to bind transiently to the same site rnay be affected by intracellular locülization of the proteins as well as binding affinities between the SH2 domains and particular phosphotyrosine residues. The Grb2 SH2 domain exhibited the strongest relative binding for p~'lm,suggesting that Grb2 would effectively out-compete Grb7 or p85 for binding to this residue. However. Grb7 also contains a PH domain (Lemmon et al., 1996) and targeting of the PH domain to the membrane in conjunction with the SH2 domain would effectively increase the amount of Grb7 at the receptor, allowing it to compete with Grb2 for binding. Furthermore, a region between the PH and SH2 domains of GrblO and Grb14 can serve as an alternate receptor binding domain (He et cd., 1998;

Kasus-Jacobi et al., 1998) and this homofogous region in Grb7 may funher contribute to

Tek binding. A similar muIti-domain scenario exists for p85 where both SH2 domains could potentially bind simultaneously to two phosphorylated tyrosine residues in the carboxy-terminal tail of Tek. Since we did not determine the binding constants for the tandem SH2 domains with either mono- or bisphosphorylated target peptides, it is possible that the paired SH2 domains would display a higher relative affinity for peptides containing pY'lWthan Grb2.

Our inability to demonstrate an in vivo interaction between p85 and Tek may suggest that the association between p85 and Tek is not direct and that p85 is actually brought to the receptor in a complex with an adaptor molecule such as Grbî, or a docking molecule such as ab2-gsociated binder 1 (Gab 1 ). However, the identification of p85 as a putative Tek binding partner in yeast suggests that this interaction is likely to occur in the absence of a bridging molecule and we were able to show that the SH2 domains of p85 cm associate with the activated Tek receptor in vitro. p85 became highly phosphorylated when coexpressed with activated Tek and this phosphorylation was abrogated following disruption of the Y"" docking site. Importantly, we have also demonstrated that p85 becomes tyrosine phosphorylated following Ang 1 stimulation of

Tek. Moreover, it has been reported that activation of a chimeric Tek receptor irz vivo results in downstream activation of PI 3-kinase and this activation is dependent on an intact Y1lmbut not an intact Y''" (Kontos et al., 1998). Collectively. these results suggest that p85 is a tme downstream signaling partner of the Angl-activated Tek receptor and its interaction is mediated through a multidocking site on Tek.

The identification of a host of signaling molecules that can associate with Tek allows us to speculate on the rote of Tek and its respective ligands in angiogenesis. For instance, Angl has recently been shown to cause endothelial cell sprouting in vitro (Koblizek et ai., 1998). This role for Angl is consistent with the findings that Tek- and

Ang 1-nu11 mice have an angiogenic or migratory defect that is manifested as a lack of vesse1 sprouting into the neuroepithelium (Sato et al., 1995; Suri et al., 1996). Sprouting angiogenesis requires migration and proliferation of endothelial cells from pre-existing vessels, and in fact out data confirrns that of Witzenbichler et ai. (1998) demonstrating that Angl is chemotactic for endothelial cells. Signaling through p85 appears to mediate endothelial ceil migration in response to Angl and PI 3-kinase signal transduction pathways have previously been shown to alter the shape and migratory properties of numerous ce11 types. However, since inhibition of PI 3-kinase can only partially abrogate the chemotactic effect of Angl on endothelial cells, it suggests that additional Tek binding partners such as Dok-R and Grb7 may contribute to Ang l -mediated endothelial cell migration. In support of this, a Caenorliabditis elegans homolog of Grb7 known as

Mig- l O appears to function in pathways that modulate neuronal ce11 migration or changes in reorganization of the cytoskeleton (Manser et ai., 1997), suggesting that Grb7 may play a similar role. Since phosphorylation of both p85 and Grb7 is dependent on Y'lM. this rnultisubstrate docking site may play a critical roie in Tek-mediated endothelial cell sprouting.

Our results further demonstrate a role for Angl in mediating endothelial ceIl survival and this effect also appears to be dependent in part on intact PI 3-kinase signal transduction pathways. The finding that Angl is a survival factor for endothelial cells in vitro is in marked contrast to the findings of Witzenbichler et al. (1998) where they demonstrate that Angl* does not protect cells from apoptosis. One factor that may account for this discrepancy is the difference between the epitope-tagged native ligand

used in these experiments and the chimeric ligand Angl* (Koblizek er al., 1998) used in

the previous experiments. PI 3-kinase has recently been shown to play a role in

promoting ce11 survival through regulation of the PH domain-containing serindthreonine

kinase PKB/Akt (Downward, 1998). In fact, activation of a chimeric Tek receptor in vivo results in downstrearn activation of PKB/Akt (Kontos et al., 1998) and Angl has ken shown to protect cultured endothelial cells from apoptosis (Kwak et al., 1999). These results collectively suggest that Angl stimulation can activate PI 3-kinase which resuits in induction of anti-apoptotic pathways that are controlled by PKB/Akt. The role of Tek in endothelial ce11 survival is consistent with the dramatic reduction in the number of endothelial cells seen in mice lacking Tek (Dumont et al., 1994). Furthermore, Tek has been shown to be constitutively phosphorylated in quiescent endothelium (Wong et al.,

1997), suggesting that constitutive activation of the PI 3-kinase signaling pathway through Tek is required for the maintenance of endothelial cells.

In addition to the unique pairs of ligands that are specific for both phosphorylation and dephosphorylation of Tek, the existence of a multifunctional docking site on Tek that appears to be required for both endothelial cell migration and survival suggests that there is exquisite control over the signaling pathways mediated by this receptor. The identification of the signaling elements and mechanisms that are controlled by Tek allows further insight into the comptex biology of endothelial cells. CHAPTER 3

THE TEWIE2 RECEPTOR SIGNALS THROUGH A NOVEL DOK-RELATED DOCKING PROTEIN, DOK-R.

A version of this chapter is published in Oncogene (N.Jones and D. J. Dumont, 1998). ABSTRACT

Tek/Tie2 is an endothelial cell-specific receptor tyrosine kinase that has been shown to play a role in vascular development of the mouse. Targeted mutagenesis of both Tek and its agonistic ligand, Angiopoietin- 1, result in embryonic lethality, demonstrating that the signal transduction pathway(s) mediated by this receptor are cruciri1 for normal embryonic development. In an attempt to identify downstrearn signaling partners of the Tek receptor, we have used the yeast two-hybrid system to identify phosphotyrosine-dependent interactions. Using this approach, we have identified a novel docking molecule calleci Dok-R, which has sequcnce and structural homology to

~62''" and IRS-3. Mapping of the phosphotyrosine-interaction domain wi thin Dok-R shows that Dok-R interacts with Tek through a PTB domain. Dok-R is coexpressed with

Tek in a number of endothelial ce11 lines. We show that coexpression of Dok-R with activated Tek resuits in tyrosine phosphorylation of Dok-R and that RasGAP and Nck coimmunoprecipitate with phosphorylated Dok-R. Furthemore. Dok-R is constitutively bound to Crk presumably through the proline rich tail of Dok-R.The cloning of Dok-R represents the first downstream substrate of the activated Tek receptor, and suggests that

Tek can signal through a multitude of pathways. INTRODUCTION

During embryonic development, the cardiovascular system is the first organ

system to form, and this is accomplished through two different processes -

vasculogenesis and angiogenesis. Vasculogenesis is the formation of discrete blood

vessels in situ from endothelial ce11 precursors, and angiogenesis is the formation of blood vessels by sprouting and extending from pre-existing vessels (reviewed in

Hanahan, 1997; Risau. 1997). Vasculogenesis is responsible for the Iarger vascular network, including the dorsal aorta and the yolk sac, while angiogenesis is required for the vascularization of organs such as the brain and kidney, as well as the remodeling of capillary networks. Angiogenesis is required in the adult for such processes as wound healing, follicular development and tumour growth.

The primary mechanisms that control vasculogenesis and angiogenesis are beginning to be eiucidated. A number of RTKs have been identified whose expression is almost exclusively restricted to cells of the endothelial lineage (reviewed in Mustonen and Alitalo, 1995). They play a crucial role in the development of the embryonic cardiovascular system, and they are also involved in angiogenesis in normal physiological and pathological processes. These RTKs are expressed in different spatial and temporal patterns during development (Dumont et al., 1995), suggesting that they play distinct roles within the endothelial cell lineage. This subgroup of RTKs has been placed into two subfamilies according to similarities in their primary amino acid sequences (Dumont et al., 1993; Mustonen and Alitalo, 1995). The first subfamily contains the high affinity vascular endothelial growth factor receptors (VEGFRs), known as FI k- l/KDlUVEGFR2, Fit- 1NEGFR 1, and Flt-4VEGFR3, and the second subfamily consists of the TEreceptors, known as Tie/TieI and TeWie2. Tie remains an orphan receptor while the ligands for Tek have recently been identified as Angl and Ang2

(Davis et al., 1996; Maisonpierre et al., 1997). Targeted nul1 mutations in all of these receptors result in embryonic lethality (Dumont er al., 1994; Dumont et al.. 1998; Fong et d.. 1995; Puri et al.. 1995: Sato et al., 1995; Shalaby et al.. I995), and the lethal phenotype is distinct for each receptor. These results suggest that the signating pathways that are mediated by these receptors play unique roles in developrnent.

The Tek receptor is first expressed around day 7.5 during mouse development, and it is expressed by endothelial precursors and by the endothelium of actively growing blood vessels (Dumont et al., 1992; Sato et al., 1993). Tek has recently been shown to be expressed at relatively high Ievels in the quiescent vasculature of a nurnber of adult tissues as well as in healing skin wounds (Wong et al., 1997). Interestingly, Tek immunoprecipitated from these quiescent and angiogenic tissues was found to be tyrosine phosphorylated, suggesting a dual role for this receptor in adult tissues. Mutational analysis of the tek locus by gene targeting has demonstrated that Tek is required for the maintenance of endothelial cells, since nul1 mice exhibit a dramatic reduction in the number of endothelial cells, and they die around embryonic day 9.5-10.5 due to malformations of the vascular network (Dumont et al., 1994; Sato et al., 1995).

Furthemore, overexpression of activated Tek in the adult leads to a hyperproliferation of endothelial cells which results in venous malformations (Vikkula et al., 1996). In this instance, when the proliferation is not coupled with the growth of smooth muscle cells, abnormal vesse1 sprouting and remodeling can occur which suggests that Tek plays a critical role in endothelial cell-smooth muscle ce11 communication during venous morphogenesis.

The recently cloned agonistic ligand for Tek, Angl, is expressed in a cornplernentary pattern to that of the receptor, suggesting a close paracrine relationship between the receptor and its ligand (Davis et al.. 1996). Interestingly, mice engineered to lack Ang 1 display angiogenic deficits that are strikingly similar to those observed in mice lacking Tek, and they also result in ernbryonic lethality (Suri et al., 1996). In contrast to

Angl. a second angiopoietin named Ang2 has been isolated. Ang2 possesses an antagonistic activity towards Tek (Maisonpierre et al., 1997) and transgenic mice engineered to overexpress Ang2 under the control of the Tek promoter also have vascular defects which resemble the Tek nuIl mutant. These results confirrn the prediction that signaling pathways downstream of Tek are required for normal vascular developrnent, and that any alterations in the signaiing pathway result in venous malformations or embryonic lethality.

Signaling by RTKs involves ligand binding followed by receptor dimerization and phosphorylation on tyrosine residues (reviewed in van der Geer et al., 1994). This autophosphorylation creates high affinity binding sites for specific intracellular signaling molecules which contain motifs that recognize phosphotyrosine, namely SH2 and PTE3 domains (reviewed in Pawson, 1995). These intracellular molecules may also serve as substrates for the autophosphorylated receptors. It is assumed that Tek will also transduce its signals by phosphorylating tyrosine residues of intracellular signaling molecules since mice engineered to lack Tek kinase activity produce the same phenotype as mice lacking the entire Tek protein (Dumont et al., 1994; Koblizek et al.. 1997).

Since the ligands for Tek have only recently been identified, there is little evidence contributing to Our understanding of Tek-mediated signding pathways. Huang rl u1. ( 1995) have identified two signding molecules, Grb2 and Shp2. that could associate with phosphorylated Tek in vitro. In order to further dissect the Tek signaling pathway in the absence of Angiopoietin-1, we have used the yeast two-hybrid system to identify novel proteins that interact with the cytoplasmic domain of the activated murine Tek receptor. We identified a novel docking molecule that shows homology to the recently cloned p62"*' and IRS-3. We tenned this related protein Dok-R and show that it interacts with Tek in an activation-dependent manner through a PTB domain both in vitro and in vivo. We also show that Tek can use Dok-R as a substrate both in vitro and in vivo and that once Dok-R is phosphorylated, it can interact with RasGAP and Nck. Furthemore,

Dok-R is constitutively bound to Crk regardless of its phosphorylation state. Our results indicate that, using an overexpression system to activate Tek, signaling from Tek can be amplified by a novei docking molecule, Dok-R.

MATEFUALS AND METHODS cDNA iibrary construction and yeast îwo-hybrid screening

Yeast two-hybrid screening was perforrned using an oligo(dT)-primed cDNA library from murine embryonic day 12.5 heart and lung tissues, prepared using the HybriZAP Two-Hybrid cDNA Gigapack CIoning Kit (Stratagene). This was sequentially

transformed into the LAO yeast strain expressing the cytoplasmic domain of Tek (amino

acids 780 to 1122) in a modified pBTM116 vector using the lithium acetate method of

transformation (Gietz et al.. 1992). Approximately 107 transfomants were plated on

synthetic medium lacking tryptophan, histidine, uracil. leucine and lysine, supplemented with 20 rnM 3-aminotriazole, and incubated at 30°C for 5 days. Hl's' transforrnants were assayed for P-galactosidase activity and putative interacting cDNAs were isolated by electroporation of yeast DNA into competent MH6 E. coli. These cDNAs were then tested to exclude any non-specific interactions by cotransformation of yeast with either pBTM116 (no insert) or LexA-Lamin. Putative interacting cDNAs that were not excluded at this stage were then tested for phosphotyrosine-dependence by cotransforrnation with either wildtype or kinase inactive Tek. cDNAs that showed blue colour specifically with the wildtype receptor were sequenced and the iden tity of each sequence revealed by the on-line BLAST search engine. Foliowing isolation of initial

Dok-R cDNA, further clones were isolated from a murine adult spleen cDNA library

(Stratagene) by standard DNA hybridization techniques.

North ern analysis

A mouse mu1 tiple tissue Northem blot (Ciontech) was hybridized under conditions recommended by the manufacturer. Blots were hybridized with various fragments from

Dok-R cDNA labeled by random primer extension (Pharmacia Biotech) using a-"P- dCTP (Dupont NEN). Immunohistochemistry

Embryos were obtained from ICR (Taconic) matings and the day of the vaginal plug was considered day 0.5. Embryos were isolated and fixed for 4 h at room temperature in

Histochoice (Amresco) followed by floating in 30% sucrose overnight at 4°C. Ernbryos were then embedded in OCT (Tissue Tek, MILES) and 10 pm thin sections were cut using a cryostat (Leica). Sections were placed on superfrost/plus slides (Fisher).

Sections were fixed for 20 min in ice cold acetone/methanol ( 1 : 1 v/v). Purified Dok-R antibody was used at 5 pg/ml and anti-platelet and endothelial ce11 adhesion rnolecule

(PECAM)- 1 (MEC 13.3. a kind gift of A. Vecchi, Milan, Italy) was diluted 1 :2000 from conditioned medium. The presence of Dok-R was visuatized with FITC anti-rabbit secondary antibody (Jackson Irnmunologicals) and the presence of PECAM-1 was determined by Cy3 anti-rat secondary antibody (Jackson Immunologicals). Slides were analyzed on a Leica DMR compound microscope.

Production of GST fusion proteins

The GST-Dok-R-PTB fusion protein was made by cloning amino acids 150-253 of mouse Dok-R into the pGEX-4T- 1 vector (Pharmacia Biotech). The GST-Dok-R fusion protein used in the kinase assay and for antibody production was made by cloning amino acids 14 1412, the region first identified in yeast. into the pGEX-4T- 1 vector. The GST fusion proteins were expressed in the DH5a strain of E. coli. GST fusion proteins were prepared using standard procedures and the recombinant fusion proteins were purified following immobilization on ghtathione-sepharose beads (Pharmacia Biotech). Purified proteins were analyzed by SDS-PAGE followed by Coomassie Blue staining. The concentrations of the proteins were estimated by cornparison with BSA standards.

GST binding and coimmunoprecip~tionassays

The cDNAs representing the cytoplasmic domain of Tek (amino acids 780 to 1 122), both wildtype and kinase inactive, were subcloned into pACTag2 (a kind gift of Jane

McGlade) to generate HA-tagged proteins. The cDNA representing the full-length Tek receptor was cloned into the pRc/RSV vector (Invitrogen) and the corresponding full- length ~ek""' cDNA was cloned into pcDNA3.1(+) (Invitrogen). 10 pg of each DNA waused to transfect a 10 cm culture dish of 293T cells, using lipofectin reagent (Gibco

BRL), according to the manufacturer's instructions. 48 h post-transfection. cells were solubilized in PLC lysis buffer [50 mM HEPES (pH 73, f 50 mM NaCi. 10% glyceroI.

1% Triton X- 100, 1 -5 mM MgC12, I rnM EGTA, 10 rnM sodium pyrophosphate, l OOmM sodium fluoride, supplemented with 1 mM sodium orthovanadate and Complete protease inhibitor tablets (Boehringer-Mannheim)], and cleared by centrifugation at 15,000 X g for 15 min at 4°C. Cleared extracts were precipitated with 4 pg of anti-Tek antibody (C-

20; Santa Cmz Biotechnology ), affinity purified anti-Dok-R antibody or purified rabbit anti-rnouse IgG (Upstate Biotechnology, Inc.) and recovered using protein A-sepharose beads (Sigma), or incubated with 5 pg of purified GST fusion protein. The precipitated proteins were eluted in twice concentrated sodium dodecyl ~Iphate(SDS) sample buffer and boiled for IO min, separated using SDS-PAGE and transferred to nitrocellulose filter using a BioRad semi-dry transfer apparatus, according to the manufacturer's instructions.

Prior to immunoblotting, membranes were bloçked with either 5% nonfat milk in TBST buffer [IOmM Tris (pH 7.5), 150 mM NaCl, and O. 1 % Tween 201 or 5% BSA in TBST

(for 4GlO blotting). Immobilized proteins were subjected to immunoblot analysis with

various antibodies. Following peroxidase-conjugated secondary antibody incubations

(BioRad) and extensive washing, membranes were developed using a chemiluminescent

reaction (ECL, Amersharn Corporation).

Peptide association assays

Unstimulated NIH 3T3 fibroblasts were lysed in PLC lysis buffer and incubated with 10

pg of biotinylated peptide for 2 h at 4°C. Complexes were recovered on streptavidin

agarose beads (Pierce), eluted in sample buffer and processed as described above.

In vitro kinase assay

Kinase assays were perforrned as described by Ziegler et al. (1993) in the presence of 1 pg of either GST or GST-Dok-R fusion protein. The reactions were allowed to proceed

for 15 min at 37°C- and they were terminated by addition of sample buffer and boiling for

1 O min. The reactions were diiuted to O. 1 % SDS in PLC lysis buffer and precleared with

protein A-sepharose. Cleared lysates were then immunoprecipitated with anti-Dok-R

antibodies (which also contained a significant amount of anti-GST antibodies) and the

immunocomplexes resolved on SDS-PAGE. Gels were dried and exposed to X-ray film.

Anfibodies used for western blotting

Commercially available antibodies used were as follows: polyclonal anti-Tek C-20

antibody (Santa Cruz); monoclonal an ti-phosphotyrosine antibody 4G 10 (Upstate Biotechnology, Inc.); monoclonal anti-HA-HRP clone 12CA5 (Boehringer Mannheim); monocional anti-Crk ôntibody (Transduction Laboratories); monoclonal anti-Nck anti body (Transduction Laboratories); monoclonal anti-RasGAP antibody (Transduction

Laboratories). Monoclonal and polyclonal anti-Tek antibodies specific to the extracellular domain were a kind gift of Fu-Kuen Lin (AMGEN, Thousand Oaks,

California). The GST-Dok-R fusion protein was used to raisc a polyclonal anti-Dok-R antiserum in rabbits and this antisemm was affinity purified. A carboxy-terminal peptide

(HIN-CQQDSSVPDWPQATEYDNVILKKGPK-COOH)was used to raise a second polyclonal anti-Dok-R antiserum in rabbits and this antiserum was also affinity purified.

This antibody is referred to as anti-Dok-Rn.

Cell lines and culture

Mouse Py4- 1 endothelial celIs have been described previousl y (Dubois et crl., 1 99 1 ).

Mouse SVR pancreatic endorhelial cells were obtained from ATCC while HEK 293T and

NIH 3T3 cells were a gift from. lane McGlade (AMGEN Institute). AI1 ceIl lines were

barown on 10 cm plates in DMEM (Gibco BRL) supplemented with 10% FBS (Gibco

BRL), 1% penicillin, 1% streptomycin, and 200 mM L-glutamine in a 37"C, 5% CO2 incubator. Ceils were grown to confluence, washed twice in cold 1 x PBS (Gibco BRL), and solubilized in PLC lysis buffer. RESULTS idenizflcation of a novef Tek binding ptoîein, Dok-R

As an approach to identify molecules that are involved in Tek signaling pathways,

we have used the yeast two-hybrid system to find proteins that can interact with the

intracellular domain of Tek in a phosphotyrosine-dependent manner (Chien et al., 199 1;

Fields and Song, 1989; O'Neill et al., 1994; Pandey et al., 1994). The entire intracellular domain of the murine receptor (amino acids 780 to 1 122) (~ek")was fused in-frame to the DNA binding dornain of the E. coli LexA transcriptional activator and this resulted in constitutive tyrosine phosphorylation of the fusion protein (Figure 3.1A). In order to screen for phosphotyrosine-dependent interactions, another fusion protein encoding a catalytically inactive forrn of Tek was generated through site-directed mutagenesis of the conserved ATP binding site of the kinase domain (Dumont et al., 1994). This mutant is herein denoted TekAss3and the corresponding intracel lular portion is referred to as

The activated receptor was used to screen for interacting proteins expressed from cDNAs obtained from a day 12.5 rnouse embryonic heart and lung library. Several clones were identified which could interact with Tek in a phosphotyrosine-dependent rnanner and one of these clones represented a portion of a novel cDNA (Figure 3.1 B).

Additional cDNA sequence was obtained by screening a murine adult spleen cDNA library and by 5' RACE on cDNAs detived from spleen, heart and kidney tissues.

Sequence analysis of several overlapping clones revealed a putative initiator methionine embedded within a good Kozak consensus (Kozak, 1986) followed by a continuous open reading frame and a poly(A) tract (Figure 3. IC). Conceptual translation Western: aTek

Figure 3.1: Expression of Tek in yeast and cloning of Dok-R. (A) Whole ce11 yeast lysates were analyzed for Tek expression and tyrosine phosphorylation using anti-Tek and anti-phosphotyrosine (pY) antibodies. ~ek".wildtype Tek intracellular domain: Tek A853Ic , kinase inactive Tek intracellular domain. (B) Growth of yeast on selective medium was dependent upon expression of an active Tek kinase. Dok-RAPHrepresents the region of Dok-R that was first identified in yeast. (C) Full-length cDNA sequence and putative open reading frame of Dok-R. The PH and PTB domains are highlighted, the 8 proline nch motifs are underlined, and the 13 tyrosine residues are in bold type. The Genbank accession number is AFO59583. (D) Deletion analysis of Dok-R in yeast defined the phosphotyrosine binding site to approximately 130 amino acids. The "+/-" indicate relative intensity of blue X-gal staining. (E) The PTB domain of Dok-R delineated in yeast shows sequence similanty to mDok, rIRS-3, and mIRS-1. The boxes indicate conserved arnino acids in al1 four PTB domains and the asterisks indicate the position of the cri tical arginine residues. AGAGAGAGGGCAGAGGCAATGGCAGTGGGGAGTGGGGAGCTGAAGCTGCGATGGTCAGGATGGAGGA MVRMEE GCCAGCTGTGAAGCAGGGCfiCCTGCATCTTCAGCAGCAGCAGACCTfiGGCAAGAAGTG PAVKQGPLHLQQQQTFGKKW GCGCCGGTTCGCAGCCGTOTTATATGGAGAGTCTGGCTGTGCCCTAGCCAGACTAGAGCT RRFAAVLTGESGCALARLEL CCAGGA~TCCCCGAGAAGACACGGCGAGGAGAGGCCACTCGGAAGGTM;TCCGCCTCAG QDVPEKTRRGEATRKVVRLS TGACTGCTTAC~AGCAGAGGTAGGCAGTGAGGCCAGCAGCCCCCGGGACACCAGTGC DCLRVAEVGSEASSPRDTSA CTTCATCCn;GAGACCAAGGAGCGCCTGTACCTACTGGCAGCCCCCTCGGCAGAGCGCAG FILETKERLTLLAAPSAERS TGACTGGATACAGGCCATCTGCC~'IYIY;GCTTTCCCGGGACA~GGAAAGGGTCACCAGG D W 1 Q À.1-C--.L L A F-P G Q R K G S P G ACTGGAGGAIVlAGAGCGGCAGTCCCTGCATGGAGGAGAACTCCAC LEEKSGSPCMEENELYSSST

601 PTB

TVLPRPESPYSRPHDSLPSP ATCCCCTGGCACACTGGTGCCTGGCATGAGGCCAGGGGCCCCTGAGGGGGAGTATGCCGT SPGTLVPGMRPGAPEGEYAV ACCCTTTGATACGGn;GCTCACTCCCTGAGGAAGAGCTTCAm-CCTCCTGACmCCC PFDTVAHSLRKSFRGLLTGP CCCTCCACACCTTCCCGACCCACTGTATGACAGCATTCAGGAGGATCCTGGGGCCCCTCT PPHLPDPLYDSIQEDPGAPL ACCTGACCACATATATGATGAGCCTGAGCCTGAGGGTGTGGCTGCCCTGTCCCTCTATGACAGGAC PRR lo81 PDHIYDEPEGVAALSLYDRT ACAGAGGCCCTCAGGGGAGACATGGAGGGAGCAGGCCACTGCCGATGGGGGTCCCAGCTC QRPSGETWREQATADGGPSS CCTCCAGCAAGACTCCTCTGTGCCTGCCTGACTGGCCACAGGCAACTGAGTATGACAATGTCAT LQQDSSVPDWPQATEYDNVI ACTTAAAAAAGGCCCAAAGTAGGGGATGAG LKKGPKt GTCTTCGGAAGCTGTGGAGGGGAAATGGTGGCTTCTffiTAGAGTCCTTTTK-TTACTGCC CCTTCCTGGGCCCCAGGCTCGGTGTGTGCCTTATTTGTGTACTGTGTGAGTCFFCTAGCCTGG AGCAGAAAGAGGGCCCTGGAGGGACCCACCCTCCTCCTACACTTTTCCACTTCCTCCCAC TGCCTTGGGTCTAAAGGGAACTGGGGCCATTTCTCTTAAAGA TGTCCTGGCTCCTCGGTGTGGGCAGGCCTCCTGCCTATAGATGTTTTGTTCCTGGGACAA CCCTGGCTTCTGGTTAGCAGCGAGAGTAGCTGTTGTCCCTCTAGCTCCCAGAGTTTTCTA TACATTAAACCAATTTCGAACTACAAAAAAAAAAA 1715 (Y- Z

fi

.PIE.i? C f f of this cDNA predicts a 412 amino acid protein with a relative molecular mass of 45K. A

search for similar sequences within GenBank revealed that this cDNA has 40 percent

similarity at the amino acid level to the recently identified RasGAP binding protein,

p62dJL(Carpino et al., 1997; Yamanashi and Baltimore, 1997) and hence we have called

this cDNA Dok-R, for Dok-Related. p62d"kand Dok-R also show structural homoiogy to

the newly cloned IRS-3 adaptor molecule (Lavan et al., 1997). Functional deletion

analysis of the Dok-R cDNA demonstrated that a small region encompassing amino acids

141 to 271 is sufficient for binding to Tek" in yeast (Figure 3.1D). This region shows

sequence conservation with the minimal domain required for phosphotyrosine binding in

IRS- 1 (Gustafson et al., 1995; O'Neill et al., 1994) and it also possesses the critical

arginine residues thought to contact phosphotyrosine in the IRS-family PTB domains

(Eck et al., 1996; Zhou et al., 1996) (Figure 3.1E). The PTT3 domains of Shc and IRS-1 . have been shown to bind to NPXY motifs in target proteins (Gustafson et al., 1995;

Kavanaugh et al., 1995; Wolf et al., 1995): however, the Tek receptor does not contain such a sequence motif suggesting that the PTB domain of Dok-R may recognize a different target amino acid sequence.

The PH and PTB domains of Dok and Dok-R share the greatest amount of sequence similarity at 48 and 58 percent. respectively (Figure 3.1E and not shown). The sequences following the PTB domain have only limited similarity throughout and these sequences are rich in proline and tyrosine residues. Dok-R contains a total of 13 tyrosine residues and many of these tyrosine residues are embedded within the appropriate context to potentially serve as docking sites for the SH2 domains of RasGAP, Crk, Abl, Src, Shp2, Nck and others (reviewed in Zhou and Cantley, 1995). Additionally, a number of the proline residues are found in a PXXP context, suggesting that Dok-R cm also interact with SH3 domain-containing proteins (reviewed in Pawson and Scott, 1997). Taken together, the overall composition of Dok-R suggests that it can interact with a wide array of sisnaling moiecules, and it could potentially mediate signaling from the activated Tek receptor to a number of downstream targets in endothelial celis, similar to IRS-1 in insulin signaling (reviewed in Yenush and White, 1997).

Dok-R is CO-expressedin different tissues and cell lines with Tek

Northem blot analysis of RNA taken from adult mouse tissues showed that Dok-R is found primarily in heart, spleen and lung (Figure 3.SA). It is expressed weakly in the liver. skeletal muscle, kidney and testis, and there is no detectable expression in the brain.

There are two predominant transcripts of approximately 2.0 and 3.0 kb in al1 positive tissues, and a third minor transcript of 4.0 kb in the kidney and testis. The complete reading frame for Dok-R can be accounted for in the 2.0 kb transcript, such that the other two transcripts may represent alternative splicing or a related gene. To address the issue of a related gene, we reprobed the blot with a probe derived from the 3' untranslated region of Dok-R and this gave an identical expression profile, suggesting that the three transcripts are indeed derived from the same gene (data not shown).

In order to study the signaling properties of Dok-R,we generated two different antibodies against Dok-R. The first antibody was raised against a fusion protein between

GST and the entire Dok-R protein without the PH domain (aDok-R) and the second Probe: 4.4-

- Western: a ~ok-R aook-fiT

Figure 3.2: Expression of Dok-R. (A) Northern blot analysis of Dok-R expression in adult mouse tissues (asterisks denote location of 3 different transcripts). P-actin expression indicates relative mRNA levels. (B) Whole ce11 lysâte preparations were made from various tissue and endothelial ce11 lines as well as from 2931 cells transiently expressing Dok-R (cDNA) and immunoblotted with aDok-R or a~ok-RCT.River, Fheart & Lung: E12.5 fetal liver and heart and lung, respectively. (C) Indirect immunofluorescence was performed on mid-sagittal thin sections cut from day 15.5 embryos. A high magnification of the heart region is shown. Low level Dok-R expression (green fluorescence) can be seen throughout the myocardium (Myo) and the trabeculae (T). The highlighted area is further magnified in (b). Expression of the endothelial marker PECAM- 1 (red fluorescence) can be detected in the endothelium in (a') and in the magnified region in (b'). Overlap of Dok-R and PECAM-1 expression can be visualized by orange fluorescence in endothelial cells in (b+b'). antibody was raised against a synthetic peptide corresponding to the last 25 amino acids of Dok-R (aDok-~~~).These antibodies were used to detect the Dok-R protein in lysates from HEK293T cells transiently transfected with Dok-R cDNA. Western blotting results show that Dok-R migrates at 56K, which is well above its predicted size (Figure 3.2B).

This aberrant gel mobility was also reported for p62&lkand it is likely due to the high proline content of these proteins. The specificity of the antisera wris confirmed since there were no reactive species detected with pre-immune sera (data not shown). Whole ce11 lysates prepared from various tissues and endothelial ceIl lines were resolved and imrnunoblotted with the sarne antibodies (Figure 3.2B). These results show that endogenous Dok-R also migrates at 56K, demonstrating that our cDNA represents the full coding region of Dok-R. These antibodies also recognize a larger protein of 72K and it may represent another isoform of Dok-R. perhaps the result of the 3.0 kb transcript. or it mriy be a rclated protein that is also recognized by Our anti-Dok-R antibodies. In summary, these results show that Dok-R can be found in a number of tissues that are rich in endothelial cells as well as in a number of cultured endothelial cell lines that have been shown to express Tek (Dumont er al., 1993) (SVR data not shown).

To further demonstrate that Tek and Dok-R are coexpressed within endothelial cells. we used affinity purified Dok-R antibody in conjunction with a marker for endothelial cells, platelet and endotheiial ce11 adhesion molecule- 1 (PECAM- I ) (MEC

13.3) (Vecchi et al.. 1994), in coimmunofluorescence studies on thin sections taken from day 15.5 embryos. Figure 3.2C illustrates that expression of Dok-R, which can be visual ized by green fluorescence (FITC) in the developing myocardium and trabeculae of the hem in panels (a) and (b), overlaps in part with the expression of PECAM-1, which is visualized by red fluorescence (Cy3) in panels (a') and (b'). Overlapping expression patterns can be visualized as orange fluorescence (FITC + Cy3) in panet (b+b'), demonstrating that Dok-R is expressed in endothelial cells and that expression of Dok-R is not restricted to endothelial cells.

The PTB domain of Dok-R mediates binding to phosphorylated Tek

The deletion analysis performed in yeast suggests that it is the PTB domain of

Dok-R that mediates binding to Tek. In order to determine whether this region can mediate binding outside of the yeast environment, we fused the PTB domain of Dok-R to

GST and purified this fusion protein from E. coli. Immobilized GST-Dok-R-PTB was incubated with lysates from S93T cells that transiently express the HA-tagged cytoplasmic domains of either a constitutively phosphorylated form of Tek (TekIc) or a kinase inactive, non-phosphory lated form (Tekhd5"C)(Figure 3.3A). Overexpression of these truncated proteins allows for the study of activated Tek in the absence of Angl.

Proteins bound to GST-Dok-R-PTB were immunoblotted for the presence of Tek. Figure

3.3B demonstrates that ~ek"can be precipitated from cell lysates by the PTB domain of

Dok-R and that this association is dependent upon the phosphorylation state of Tek.

These results demonstrate that the Dok-R PTB domain is able to bind specifically to Tek irr vitro and that this interaction is Tek tyrosine kinase activation-dependent. A B Western : aHA wrrr

IP: Tek GST-PTB GST

**O-97- !lmll!P - 52 - Western : Tek PY Tek -Dok-R - IP: Tek Ook-R IgG WCL

Figure 3.3: In vitro and in vivo association of Tek with Dok-R. (A) Lysates from 293T cells expressing a tagged, tmncated form of Tek (~ek") or kinase inactive Tek (~ek*''~") were immunoprecipitated with anti-Tek C-20 antibody. (B) Lysates from (A) were incubated with irnmobilized GST-Dok-R-PTB (GST-PTB)or GST alone. Resulting complexes were resolved by SDS-PAGE and immunoblotted with anti-HA-HRP. (C) Full-length Tek or ~ek*"~were transiently coexpressed with Dok-R in 2931 cells. Lysates were immunoprecipitated with either a polyclonal anti-Tek antibody specific to the extracelluIar domain of Tek or anti-Dok-R antibodies or purified rabbit anti-mouse IgG antibodies, resolved by SDS-PAGE, and immunoblotted as indicated. Non- immunoprecipitated lysates (whole ce11 lysate-WCL) show equal amounts of Dok-R in the initial lysates following immunoblotting with anti-Dok-R antibodies. In vivo associalion of Dok-R and Tek in 2937 cells

To further determine whether the association we detected in yeast and in vitro could occur in vivo, we transiently expressed either Tek or TekA"' with Dok-R in 293T cells. Overexpression of full-Iength Tek in 293T cells results in its tyrosine phosphorylation (Figure 3.3C). This phosphoryiation is dependent upon a functional Tek kinase domain suggesting that the sites that are phosphorylated under these conditions are defined by the specificity of the Tek kinase domain and as such should reflect the sites that are normally phosphorylated upon ligand stimulation. However, we cannot rule out the possibility that this pattern of phosphorylation would not be recapitulated under conditions of Ang 1 stimulation in vivo- Lysates were prepared from these transfected cells and immunoprecipitated with an antibody specific for Dok-R. Coprecipitation of

Tek with Dok-R was only observed when Tek was phosphorylated, dernonstrating that

Dok-R crin associate with Tek in vivo and that the interaction is dependent upon Tek kinase activity (Figure 3.3C).

Tek can phosphorylate Dok-R in vitro and in vivo

A number of receptors once activated will utilize the proteins bound to them as substrates. For instance, when the insulin receptor is stimulated by insulin, it specifically recruits the docking molecule IRS-1 through its PTB domain and subsequently phosphorylates IRS-1 on tyrosine residues (Sun et al., 1993). In addition, the EGF receptor (EGFR) can use the adaptor molecules Shc and Gabi as substrates when stimulated (Holgado-Madmga et al., 1996; Pelicci er al., 1992). To demonstrate that Tek could utilize Dok-R as a substrtite, we immunoprecipitated full-length activated Tek from Tek-transfected 293T ceIl lysates and perforrned in vitro kinase assays using either GST or GST-Dok-R as substrates. A Tek immunoprecipitate was able to phosphorylate the recombinant GST-Dok-R whereas no phosphorylation of GST alone was detected (Figure

3.4A).

To determine whether Tek could utilize Dok-R as an in vivo substrate, we immunoprecipitated Dok-R from celis transfected with either Dok-R alone or in combination with Tek or ~ek"'". Proteins from these immunoprecipitates were resolved and immunoblotted for the presence of phosphotyrosine. Figure 3.4B demonstrates that

Dok-R is only tyrosine phosphorylated in the presence of a kinase active form of Tek.

Dok-R coimmunoprecipitation with the phosphorylated receptor can aIso be seen here.

Although we cannot rule out the possibililty that activation of Tek results in the stepwise activation of another kinase which is responsible for the phosphorylation of Dok-R, our in vitro and in vivo data suggests that upon recruitment of Dok-R to phosphorylated Tek,

Tek is able to utilize Dok-R as a substrate and subsequently cause its tyrosine phosphoryiation.

PItospIrorylation of Dok-R establishes binding sites for RasGAP and Nck

Several receptor binding proteins Iike IRS-1 and Shc that are phosphorylated by the receptor function as docking molecules for other cytosolic SH2- and PTB-containing proteins (reviewed in Pawson and Scott, 1997). This phosphorylation by the receptor establishes high affinity binding sites for these proteins. Scanning the sequence of Dok-R for potential binding sites revealed that Dok-R contains four potential RasGAP. Crk and Western: pY Dok-R IP: Dok-R WCL

Western :

Western: Nck t Nck

Figure 3.4: Dok-R becomes tyrosine phosphorylated by Tek in vitro and in vivo and it can serve as a docking molecule. (A) Immunoprecipitated Tek can utilize GST-Dok-R but not GST alone as a substrate in an in vitro kinase assay. (B) 293T cells were transfected with Dok-R in combination with Tek or TekA853 . Lysates were immunoprecipitated with anti-Dok-R antibodies, resolved by SDS-PAGE, and immunoblotted with anti-phosphotyrosine antibodies. Dok-R is indicated by the arrow and Tek is indicated by the asterisk. Equal amounts of Dok-R are present in the initial I ysates. (C) Lysates from (B) were immunoprecipitated with anti-Dok-R antibodies and immunoblotted as indicated. The arrows indicate the position of the proteins consistent with control lysates. (D) A predicted Nck binding site is conserved in Dok-R and ~62~"'. A synthetic phosphopeptide corresponding to this region of Dok-R (pY351) was able to precipitate Nck from fibroblast lysates whereas an unphosphorylated peptide (Y35 1) was not. WCL, whole ce11 Iysate: MDCK,Madin-Darby canine kidney epithelial cells. Nck binding sites (YXXP) (Songyang et al., 1993). To determine whether phosphorylation by Tek establishes binding sites for these SH2-containing proteins in vivo, we expressed either Dok-R alone or in combination with Tek or TekAg5' in 293T celis. Lysates from these transfected cells were immunoprecipitated with an antibody specific for Dok-R,and the resulting complexes were resolved and immunoblotted with various antibodies. These results show that both RasGAP and Nck can coi mmunoprecipitate with Dok-R in a phosphory lation-dependent manner, while Crk can coirnmunoprecipitate with Dok-R regardless of its phosphory lation state (Figure 3.4C).

These results suggest that the interaction of both RasGAP and Nck with Dok-R is likely rnediated by their SH2 domains with phosphotyrosine residues on Dok-R, while the interaction between Crk and Dok-R is likely constitutive through one of the SH3 domains of Crk and a proline rich sequence of Dok-R. Interestingly, Dok and Dok-R show striking homology in a small area of their carboxyl region (Figure 3.4D). In fact, this homology is conserved in both human and murine Dok and Dok-R proteins (not shown), suggesting that this region plays an important role in their signaling. The tyrosine residue in this region is found within a predicted SH2-binding site for Nck (YDEP). Using phosphorylated and unphosphorylated biotinylated peptides corresponding to the region surrounding tyrosine residue 351 (Y351)of Dok-R, we demonstrated that the phosphorylated peptide derived from this highly conserved region will mediate the binding of Nck to Dok-R while the unphosphorylated peptide will not (Figure 3.4D).

This result suggests that signaling through Nck is likely a common feature in this newly emerging family of docking molecules. Taken together, these results show that phosphorylation of Dok-R by Tek enables it to function as a docking molecule during

Tek signaling through its interactions with a number of downstrearn signaling molecules.

DISCUSSION

In this study, we describe the identification of a novel docking molecule, Dok-R, which can interact in a phosphorylation-dependent manner with the endothelial receptor,

Tek. The binding of Dok-R to Tek both in yeast and in virro was mediated through a region of Dok-R that shows homology to the PTB domains of Dok and IRS-1. Dok-R also contains an arnino-terminal PH dornain and this is likely involved in recruitment of

Dok-R to the ce11 membrane. Dok-R becomes tyrosine phosphorylated in the presence of an activated Tek receptor, and this is the first reported substrate of the Tek receptor.

Phosphorylation of Dok-R establishes binding sites for RasGAP and Nck. and Dok-R is constitutively associated with Crk. In surnmary, these results present evidence to suggest that Dok-R functions as a novel docking molecule that can link the activated Tek receptor to a number of different signaiing pathways.

Dok-R contains a central PTB domain that is essential for its binding to the Tek receptor in vitro. Sequence analysis of this PTB domain shows that it is related to the

PTB domain of IRS- 1, and it contains the critical arginine residues that have been shown to contact the phosphotyrosine residue in the target binding site. It is interesting to note that the Tek receptor does not contain the reported consensus target sequence, NPXY. A similar paradigm has been reported for the fibroblast growth factor (FGF) receptor and its newly defined substrate, FGF Leceptor substrate 2 (FRS2)(Kouhara et al., 1997). FRS2 contains a putative PTB domain that is also related to the FI% domain of IRS-1, and, although it is not clear whether the interaction between the FGF receptor and FRS2 is direct, the FGF receptor does not contain an NPXY motif. Additionally, the PTB domain of Shc has been shown to interact with PTP-PEST through an NPLH motif (Charest et al., 1996) and the PTB domains of XI 1 and FE65 (Borg et al., 1996) have been shown to interact with non-phosphorylated NPXY motifs. Similarly, the l'TB domain of Numb has been shown to interact with non-phosphorylated (Dho et al., 1998) and non-NPXY containing motifs (Chien et al., 1998; Li et al.. 1997). This suggests that the specificity of PTB domain binding may be more dependent upon tertiary structure rather than on a l inear sequence.

The amino terminus of Dok-R contains a PH domain (Lemmon et al.. 1996). PH domains have been shown to mediate binding of various proteins to phosphoinositides on the inner face of the plasma membrane, thus serving as a domain to target host proteins to the membrane, presumably in the vicinity of their physiological targets. A number of docking molecules contain both a PH domain and a PTB domain, which essentially provides two contributions to membrane localization for the host protein. In the case of

IRS- 1, intact PH and PTB domains are required for efficient coupling of IRS- 1 to the insulin receptor and tyrosine phosphorylation of IRS-1 (Gustafson et al., 1995; O'Neill et cri., 1994; Yenush et al., 1996). The role of the PH domains of Dok and Dok-R is currently unknown, although they can be expected to perform a similar function in mediating phospholipid binding and subseqüent membrane targeting upon receptor stimulation. The cloning of Dok-R marks the first evidence for a downstream substrate of the activated Tek receptor. Dok-R becomes tyrosine phosphorylated in cells expressing an active Tek kinase and this phosphorylation establishes binding sites for a number of cytosolic signaling proteins that contain SH2 domains. Dok-R contains a total of 13 tyrosine residues that could potentiaily serve as docking sites. Our results show that phosphorylation of Dok-R in vivo creates docking sites for the SH2 domains of RasGAP and Nck. Additionally, we have shown that Dok-R is constitutively coupled to Crk. and this interaction is likely rnediated by one (or more) of the 8 PXXP motifs in the Dok-R protein binding to one of the SH3 dornains of Crk. The presence of this multitude of inducible SN2 and constitutive SI33 docking sites in the Dok-R protein suggests that

Dok-R can interact with a wide array of signaling andor scaffolding proteins. Our findings that Dok-R can interact with RasGAP, Nck and Crk demonstrates that Tek- mediated signaling may proceed through a number of different cellular pathways that could affect ce11 rnobility.

Nck is an SH2 and SH3 domain-containing adaptor protein that is expressed in a number of tissues and ce11 lines. The Drosophila melanogasrer homolog of Nck,

Dreadlocks, has been shown to play an important role during axonal navigation in photoreceptor ceIls (Garrity et al., 1 996), but the role of Nck in rnarnmalian cells is unclear. The SH2 dornain of Nck has previously been shown to bind to phosphorylated receptor and cytoplasmic tyrosine kinases as well as to RS- 1 (reviewed in Birge et al., f 996). In addition, the SH3 domains of Nck can constitutively couple to a nurnber of proteins including the serinehhreonine kinases NAK, PRK2, Pakl and CKI-72, the guanine nucleotide exchange factor, SOS,the protein product of the c-cbl proto-oncogene and the Wiskott-Aldrich syndrome protein, WASP. The interaction between Nck and this host of signaling molecules allows Nck to play a role during cellular proliferation and in the induction of actin cytoskeleton reorganization. The physiological role of the interaction between Nck and Dok-R is unclear at this stage. Since Nck has been shown to play a role in ce11 migration following growth factor stimulation (Rockow et al.. 1996)- it is plausible that the inducible interaction between Nck and Dok-R following Tek stimulation wiIl also affect endothelial cell migration.

Crk is an adaptor protein that is differentially spliced to contain either one or two

SH3 domains in conjunction with an SHS domain (reviewed in Matsuda and Kurata,

1996). The SH3 domains of Crk mediate interactions with two guanine nucleotide exchange proteins, known as SOSand C3G. SOSis an exchange factor for Ras which results in activation of the MAPK pathway, and Crk has been shown to be essential for positive activation of Ras (Tanaka et al.. 1993). In contrast, C3G is an activator of Rap 1, which is another Ras-family protein (Gotoh et al., 1995). Overexpression of C3G in NIH

3T3 cells cotransfected with activated Ras had a reverting effect on Ras-mediated transformation, suggesting that Rap 1 may act as a suppressor of Ras (Gotoh et al., 1995).

Additionally, Crk has recently been shown to activate c-Jun kinase (JNKl ) through C3G

(Tanaka et al., 1997), suggesting that Crk could potentially contribute to the activation of parailel signaling pathways. Binding studies with Crk have shown that the SH2 domain of Crk associates with a nurnber of proteins found in focal adhesion kinase (FAK) complexes (reviewed in Hanks and Polte, 1997). Signaling through FAK affects the actin cytoskeleton to ultimately change the shape and migratory properties of cells. Whether

Tek-signaling mediated through Dok-R affects any of these processes by virtue of its association with Crk is unknown.

One obvious connection between Dok-R and endothelial ceIl signaling can be drawn from the inducible interaction between Dok-R and RasGAP. RasGAP is a GTPase activating protein and it functions as a negative regulator of Ras signaling by stimulating the intrinsic rate of Ras GTPase activity (Trahey et al., 1988). Deletion of the RasGAP pene (Henkemeyer et al.. 1995) produces a very pleiotropic phenotype as would be expected for a key enzyme that rests downstream of multiple signaling pathways.

Vascular defects observed in RasGAP-nul1 embryos (Henkemeyer et al., 1995) resemble the loss of vascular integrity seen in both Tek- and Angl-nul1 embryos (Dumont et al.,

1994; Sato et al., 1995; Suri et al., 1996). Furthermore. an abnomal endotheli al ce11 migration pattern was observed in these embryos. Since Ras is required for the motility of cultured endothelial celis in vitro (Sosnowski et al., 1993) it is likely that this phenotype may be a result of inappropriate signaling through Ras in the endothelial cells of these embryos. The vascular defects observed in RasGAP-nul1 embryos may in part be due to a defect in the Tek signaling pathway, since the absence of RasGAP in these endothelial cells rnay alter the ability of Dok-R to affect changes in Ras activity following Tek stimulation. The interaction of Dok-R with this trio of signaling molecules suggests that Tek

signaiing through Dok-R may ultirnately lead to alterations in the intracellular architecture of endothelial cells through a number of different pathways. This change in cellular morphology could facilitate the directed motility/migration of endothelial cells during angiogenesis, which is a prerequisite for the sprouting process. This is supported by the fact that endothelial cells in Tek- and Angl-nul1 embryos are thought to have an angiogenic or migratory defect, and this is manifested as a lack of vesse1 sprouting

(Dumont et ni., 1994; Sato et al., 1995; Suri et al., 1996).

The ability of Dok-R to interact with RasGAP, Nck and Crk suggests that it functions as a multi-site docking molecule during Tek signaling. There is also evidence to suggest that Dok may be playing a similar role as a docking molecule. Carpino et al.

(1997) have shown that p62""%ecomes tyrosine phosphorylated upon c-Kit ligand stimulation of Mo7 leukemic progenitor cells and Yamanashi and Baltimore ( 1997) have shown that Dok binds to (and is phosphorylated by) v-Abl. Tyrosine phosphorylated Dok can then interact with RasGAP. Additionally, Dok was also phosphorylated upon stimulation of the EphB2 receptor (Holland et cil., 1997) and the EGFR (Tang et al.,

1997) and in both instances, phosphorylated Dok was associated with Nck. Taken together, this suggests that upon tyrosine phosphorylation, Dok and Dok-R may interact with a similar repertoire of downstream substrates. However, it has been reported that

Crk does not associate with ~62~0~(Yamanashi and Baltimore, 1997), suggesting that

Dok and Dok-R do exhibit different specificities in the molecules that they will recmit. In conclusion, the preceding results show that upon Tek activation and phosphorylation, the receptor is able to specifically recruit a novel docking molecule caIled Dok-R. Furthemore, Tek can utilize Dok-R as a substrate in vitro and when Dok-

R and Tek are coexpressed under conditions that result in the activation of Tek, Dok-R becomes tyrosine phosphorylated. This phosphorylation of Dok-R establishes high- affinity binding sites for two SH2-containing proteins RasGAP and Nck, and Dok-R is constitutiveiy bound to Crk. These experiments were accomplished in the absence of

Angl stimulation; however, the phosphorylation of Tek in our assay is solely dependent upon an active membrane-localized Tek enzyme suggesting that the phosphorylated residues are those which would nonnally be phosphorylated upon Ang 1 stimulation. The subsequent recruitment of these cytosolic proteins suggests that Tek-mediated signaling may proceed through a number of different pathways which allows for diversification and amplification of signals from Tek. The identification of Dok-R and the subsequent elucidation of Tek-mediated signaling pathways will contribute to our understanding of normal vascular development in both the embryo and the adult, as well as in pathological situations such as tumour-mediated angiogenesis. CHAPTER 4

ASSIGNMENT OF THE DOKR GENE TO MOUSE CHROMOSOME 14D2-D3 BY FLUORESCENCE IN SITU HYBRIDIZATION.

A version of this chapter is published in Genomics (N.Jones and D. J. Dumont, 1998). BRIEF MAPPING REPORT

Furtctional gene description: Signal ing through the Temie-2 endothel ial-specific receptor tyrosine kinase has been shown to play a role in normal vascular development of the mouse (Dumont et al., 1994). We have recently identified a novel docking molecule known as DOKR that can couple the activated Tek receptor to a number of downstream effector molecules, including RasGAP, Nck and Crk (Jones and Dumont, 1998a).

DOKR, which is also known as p~6~'"''(GenBank Accession Nos. AF034970 and

AF03S 1 17). shows structural and to ~62''~(hwnstream qf tyrosine

-kinases) (GenBank Accession Nos. U70987 and U788 18). The structurai features that define this emerging family of docking molecules are an amino-terminal PH domain, a central PTB domain and a carboxy-terminal proline rich region, as well as a large number of tyrosine residues. These structural domains and motifs collectively allow these docking molecules to couple with a wide may of downstream proteins.

DNA source and its description: The Dokr clone was isolated and purified from a Pl bacteriophage mouse genomic library (Genome Systems Inc., St. Louis, MO).

Method used to validnle gene identiiy: The Dokr gene was identified using gene-specific primers on individual PI genomic clones (Genome Systems Inc.). Identity of the Dokr gene was confirmed following hybridization of the DOKR cDNA with the purified Pl genomic DNA by Southem analysis. Sequence data for DOKR have been deposited with the GenBankEMBL database under Accession No. AF059583. Fùznking markers used: A Pl clone specific for the centromeric region of chromosome

14 was cohybridized with the labeled Dokr Pl clone. The chromosome 14-specific P 1 clone was identified by hybridization with labeIed probe DNA amplified from genomic

DIVA using the DlIMIT12 primer pair (Shi et al., 1997)- The identity of this clone was confirrned by fluorescence in situ hybridization (FISH) mapping.

Metlrod of mapping: Punfied Pl genornic DNA was labeled with digoxigenin-dUTP by nick-translation and hybridized to normal metaphase derived from mouse embryo fibroblast cells in the presence of 50% formamide, 10% dextran sulphate and mouse genomic DNA. Specific FISH signals were detected by incubating the hybridized slides in fluorescein-conjugated anti-digoxigenin antibodies followed by counterstaining wi th DAPI for chromosome identification.

Resuïfs: Initial experiments resulted in the specific labeling of the central portion of a relatively small-sized chromosome that was believed to be chromosome 14 on the basis of DAPI staining (Figure 4. IA). Cohybridization of the Iabeled Dokr gene with the chromosome 14 centromere probe resulted in the specific labeling of the centromere and the middle portion of chromosome 14 (Figure 4.1 B). Measurements of 10 specifically hybridized chromosomes 14 demonstrated that Dokr is Iocated at a position which is 5 1 % of the distance from the heterochromatic-euchromatic boundary to the telomere of chromosome 14, an area that corresponds to mouse band I4D2-D3 (Figure 4. IC). In these experiments, a total of 80 metaphase cells were analyzed, with 71 of these cells exhibiting positive labeling. In summq, FISH mapping demonstrates that the mouse Figure 4.1: Localization of the mouse Dokr gene by fluorescence in siru hybridization. (A. B) FISH on normal metaphase chromosome spreads derived from mouse embryo fibroblast cells. Mouse Dokr genornic Pl clone was labeled with digoxigenin-dUTP and visualized by fluorescein-conjugated anti-digoxigenin antibodies followed by DAPI counterstaining. Specific Dokr hybridization sites on chromosome 14 sister chromaiids are indicated by arrows and the chromosome 14-specific hybridization sites are indicated by arrowheads. (C) Idiogram of G-banded mouse chromosome 14: the position of the Dokr locus is indicated by the arrowhead. Dokr gene can be localized to mouse chromosome 14, band D2-D3. This region is syntenic to human chromosomes 8p (near band 8p2 1) and 14q (near band 14q 1 1).

Additional comments: Recent studies have shown that Ang2 maps to human chromosome 8p23, although there are no data to suggest any linkage groups in this region

(Cheung et al., 1998). The T-ceIl receptor alphddelta locus maps to human chromosome band 14q 1 1, and rearrangements at this locus have been detected in a variety of T- and B- lymphoproliferative disorders (Bernard et al., 1993). Additionally, loss of heterozygosity for chromosome 14q has been detected in low- and high-grade meningiomas while loss of heterozygosity for 8p has been shown to play a role in hepatocellular carcinoma.

Further studies will be required to address the function of the DOKR protein as it relates to tumourigenesis. CHAPTER 5

RECRUITMENT OF DOK-R TO THE EGF RECEPTOR THROUGH ITS PTB DOMAIN IS REQUIRED FOR A'ITENUATION OF ERK MAP KINASE ACTIVATION.

A version of this chapter is published in Current Biology (N.Jones and D. J. Dumont, 1999). ABSTRACT

Dok proteins are a newly identified family of docking molecules that are characterized by the presence of an arnino-terminal PH domain, a central putative PTE! domain and numerous potential sites of tyrosine phosphorylation (Carpino et al., 1997;

Di Cristofano et al., 1998; Jones and Dumont, 1 998a; Lock et al., 1999: Nelms et al..

1998; Yamanashi and Baltimore, 1997). Here, we explore the potential role of the Dok farnily member Dok-R (also known as ~56'"" or FRIP) in signaling pathways mediated by the EGFR. An intact PTB domain in Dok-R was critical for its association with two

PTB-binding consensus sites on the EGFR and the PH domain further contributed to stable in vivo binding and tyrosine phosphorylation of Dok-R. Multiple sites on Dok-R were tyrosine phosphorylated following EGF stimulation; phosphorylated tyrosine residues 276 and 304 are proposed to dock the tandem SH2 domains of the p2IR"'

GTPase-acti vating protein RasG AP and tyrosine residue 35 1 mediates an association with the SH2 domain of the adaptor protein Nck. Interestingiy, we have found that Dok-

R could attenuate EGF-stimulated MAPK activation independently of its association with

RasGAP. Together, these results suggest that Dok-R has an important roie downstream of growth factor receptors as a potential negative regulator of signal transduction. MATERIALS AND METHODS

Plasmids

The cDNA encoding full-length Dok-R (Jones and Dumont, 1998a) was subcloned into

pcDNA3.1 (+) (Invitrogen) or pcDNA3.1 (+) containing an HA-tag (HA-pcDNA3.1; a

kind gift of Bryan Snow, AMGEN Institute, Toronto, Canada). Point mutations were

introduced using the QuikChange site-directed mutagenesis (Stratagene) and al1

mutations were confirmed by sequencing. To generate Dok-~~~~'and Dok-R GST-

PTB*. arginine residues 202 and 2 17 were altered to glutamine. Dok-RAPH(amino acids

143-4 12) was constnicted using PCR with appropriate primers containing restriction sites

suitable for in-frame insertion into HA-pcDNA3.1. The Dok-R GST-PTE? construct was

described previously (Jones and Dumont, 1998a) and the pEBG-MAPK construct was a

gift of Jim Woodgett (Ontario Cancer Institute, Toronto, Canada).

Cell culture, transient transfections and binding assays

COS 1 cells were rnaintained and processed as described in (Jones and Dumont, 1998a) except that cells were serum-starved for approximately 24 h in DMEM containing 0.1 %

FBS (Gibco BRL) before stimulation with iûûnb/mL EGF (Upstate Biotechnogy Inc.).

Myelin basic protein (MBP) phosphorylation assays were perforrned essentially as described in (Keilhack et al., 1998).

GSTfusion proteins and antibodies

Production of GST fusion proteins has been described in (Jones and Dumont, 199th).

The SH2-SH3-SH2 domain fusion protein of RasGAP was purchased from Santa Cruz. Peptide association assays are described in (Jones et al., 1999) and the sequences corresponding to the phosphorylated Dok-R peptides are as follows: Y276-

RPESPpYSRPHDSL; Y304-APEGEpYAVPFDTV; Y35 1-DHIpYDEPEGVA; Y402-

PQATEpYDNVILKK. Affinity purified polyclonal anti-Dok-R antibodies have been

described previously (Jones and Dumont, 1998a). Polyclonal anti-Shc antibodies were a

gift of Jane McGlade (Hospital for Sick Children, Toronto, Canada). Monoclonal anti-

EGFR antibodies (E12020, Transduction Labs) in conjunction with rabbit anti-mouse

IgG (Upstate Biotechnology, Inc.) were used for immunoprecipitation and polyclonal anti-EGFR antibodies (1005, Santa Cruz) were used for Western blotting. Monoclonal anti-HA clone 12CA5 (Boehringer Mannheim) was used in immunoprecipitations and as a horseradish peroxidase conjugate for Western blotting. The remaining antibodies used for Western blotting were obtained as follows: monoclonal anti-phosphotyrosine 4G 10

(Upstate Biotechnology, Inc.); monoclonal anti-Grb2, anti-RasGAP and anti-Nck

(G 16720, G 12920 and N 15920, Transduction Laboratories): monoclonal phospho-MAPK p44/p42 (E 1 O, New England Biolabs); and polÿclonal anti-GST (2-5,Santa Cruz).

RESULTS AND DISCUSSION

Tyrosine phosphorylation of Dok in response to EGF stimulation has previously been reported (Ellis er al., 1990: Moran et al., 1991). Given the similar expression profiles of Dok and Dok-R, we reasoned that Dok-R might also function in EGF- rnediated signaling pathways. To test this, we used the COS1 ce11 line that expresses high levels of EGFR (Gluzman, 198 1) and does not contain endogenous Dok-R (data not shown); there is thus no interference with the analysis of signaling mediated by exogenously expressed forms. In COS 1 cells transiently expressing full-Iength Dok-R,

we found that EGF stimulation resulted in complex formation between the EGFR and

Dok-R and in tyrosine phosphorylation of both proteins (Figure 5.1).

A region of Dok-R similar to the PTB domain of IRS-1 has previously been

shown to mediate binding of Dok-R to the phosphorylated Ternie2 receptor (Jones and

Dumont, 1998a). As the sequence similarity between PTB domains is limited and since

the Ternie2 receptor does not contain the consensus binding site for PTE3 domains, we

wanted to ascertain whether this region of Dok-R represented a true PTB domain.

Purified fusion proteins of GST with the putative Dok-R PTB domain (GST-PTB) or a mutant domain lacking two conserved arginine residues (GST-PTB*) (Eck et al., 1996;

Zhou et al., 1996) were used to precipitate the EGFR from lysates of non-transfected unstimulated or EGF-stimulated COS 1 cells. Figure 5.2A shows that the PTE3 domain of

Dok-R could precipitate the EGFR following EGF stimulation and that the PTB* domain could not.

To determine which tyrosine residues on the EGFR could mediate the interaction with Dok-R,we designed a series of synthetic peptides corresponding to the sequences surrounding the autophosphorylation sites of the EGFR (Figure 5.2B) (Downward et al.,

1984) and mixed them in vitro with purified GST-PTE3 fusion proteins. The GST-PTB domain of Dok-R could be precipitated by phosphopeptides representing tyrosine

residues 1086 (pY 1086) and 1 148 (pY 1148) but not by the unphosphorylated counterparts (Y 1086 and Y 1 148) (Figure 5.2C). In contrast, neither GST-PTB* nor GST IP: EGFR Dok-R Lysate EGF: = + - + - +

Blot: a-pY

Figure 5.1: Dok-R is a novel substrate of the EGF receptor. Lysates from unstimulated (-) or EGF-stimulated (+) COS! cells transiently expressing full-length Dok-R were immunoprecipitated (IP) with anti-EGFR or anti-Dok-R antibodies and irnmunoblotted (Blot) as indicated. Non-immunoprecipi tated Iysates show equal amounts of EGF receptor and Dok-R in the initial lysates. A GST GST GST PT8 PTB' Lysate

I EGF: - + œ + œ + +

p;rl OC* EGFR

Blot: a-EGFR

Figure 5.2: Dok-R contains a functional PTB domain that rnediates the association with two NPXY motifs of the EGFR. (A) Lysates from unstimulated (-) or EGF- stimulated (+) COS1 cells were incubated with immobilized GST, Dok-R GST-PTB or Dok-R GST- Pm*. Bound proteins were detected with anti-EGFR anti bodies (Blot). EGF receptor could be detected in both unstimulated and EGF-stimulated non- precipitated lysates (Lysate). (B) Schematic representation of the intracellular domain of the EGF receptor showing three tyrosine autophosphorylation sites (Y 1086, Y 1 148 and Y1 173) and one non-phosphorylated site (Y 11 14). The remaining autophosphorylation sites (Y992 and Y1068) were not studied here. The synthetic biotinylated peptides corresponding to these sites that were used in the mapping studies are shown with the asparagine (N), proline (P) and tyrosine (Y)residues that were altered in similar peptides in bold. (C) Associations between the various EGF receptor peptides and purified GST-fusion proteins were allowed to form in vitro and resultant complexes were recovered using streptavidin-agarose beads. Bound fusion proteins were visualized with anti-GST antibodies. The lower two panels show different exposures of the sarne immunoblot. GST

GST-PT B

GST-PT B*

GST-Shc

GST-PT B r-- GST-PTB afone could be precipitated by either of these phosphopeptides and the PTI3 domain of

Shc (GST-Shc) associated almost exclusively with pY 1148 (Figure 5.2C). Tyrosine

residues 1086 and 1 148 are found within the predicted autophosphorylated NPXY motifs,

so to investigate the contribution of the asparagine and proline residues in target binding,

we designed synthetic phosphopeptides in which either of these residues were altered to

alanine. Mutation of either of these residues resulted in abrogation of the interactions,

although weak binding to phosphopeptides lacking the proline residues (pY 1086'~and

pY 1 14SPA) could still be seen in longer exposures of the immunoblot (Figure 5.X).

Estimates of the dissociation constants (K,) between pY 1 O86 and pY 1 148

phosphopeptides and the soluble PTE3 domain of Dok-R revealed that the PTB domain could bind to both phosphopeptides with similar relative affinities (Table 5.1 ).

Our results to this point have demonstrated that the WB domain of Dok-R is required for binding of Dok-R to the phosphorylated EGFR in vitro. As Dok-R also contains a PH domain, however, we wanted to examine the contribution of each domain to the association Nz vivo. We therefore generated a series of mutations in full-Iength

Dok-R that would disrupt the functions of the individual domains, introduced the mutants into COS 1 cells and examined their ability to associate with the EGFR and undergo tyrosine phosphorylation in response to EGF. Disruption of the PTB domain (PTB*) appeared to have a more dramatic effect on EGFR binding and Dok-R phosphorylation than deletion of the PH domain (APH) when compared to wildtype levels (WT) although the disruption of both domains completely abrogated Dok-R phosphorylation

(PTB*/APH) (Figure 5.3). TABLE 5.1

Determination of the relative affinity of the PTB domain of Dok-R for phosphopeptides representing tyrosine residues 10U and 1148 of the EGFR using BIAcore analysis.

EGF Receptor Phosphopeptide Binding Constant (K,)

Tyr 1 O86 Tyr1 148

Table 5.1: The PTB domain of Dok-R binds EGFR phosphopeptides with similar relative affinities. Phosphopeptides representing tyrosine residues 1086 (TyrIO86) and

1 148 (Tyr1 148) were immobilized on sensor chips and the dissociation constants (K,) between each of these phosphopeptides and the soluble PTB domain of Dok-R were determined using real time biosensor (BIAcore) analysis (Jones et al., 1999). This analysis provides an estimate of the relative affinity of the PTB domain for each of the two phosphopeptides. These values are likely to be overestimates because of the dimeric nature of GST in these experiments (Ladbury et al.. 1995) IP: Dok-R PTB+ -\NT -PTB" APH APH VECTOR- EGF:+-+-+++

+ EGFR

Blot: a-Dok-R

Figure 5.3: The PTB domain of Dok-R is required for binding to the EGF receptor in vivo and for tyrosine phosphoryiation. Full-length wildtype Dok-R (WT) or mutant forms of Dok-R containing inactivating mutations in the PTB domain (PTB*), a deletion of the entire PH domain (APH) or both mutations (PTB*/APH)were inuoduced into COS1 cells. Unstimulated or EGF-stimulated lysates frorn these cells were imrnunoprecipitated (IP) with anti-Dok-R antibodies and immunoblotted with anti-phosphotyrosine (pY) antibodies. Immunoblotting with anti-Dok-R antibodies shows approximately equal expression of al1 Dok-R mutants and no expression of Dok- R is detected in cells expressing the vector alone. Phosphorylation of Dok-R has previously been shown to establish docking sites for downstrearn SH2 domain-containing signaling proteins including RasGAP and Nck

(Di Cristofano et al., 1998; Jones and Dumont, 1998a; Nelrns et al., 1998) and in these expetirnents, we found that RasGAP and Nck could associate with Dok-R following EGF stimulation (Figure 5.4A). The consensus binding motif for both RasGAP and Nck is

YXXP (Holland er al.. 1997; Hu and Settleman, 1997; Songyang et al., 1993); this motif is found four times in the Dok-R amino acid sequence at tyrosine residues 194, 276, 304 and 35 1. Synthetic phosphopeptides corresponding to the last three YXXP motifs were generated; the amino-terminal SH2 domain of RasGAP (GAP-N) was found to associate with al1 three YXXP phosphopeptides in vitro, whereas the SH2 domain of Nck could associate only with phosphopeptides corresponding to tyrosine residues 304 and 35 1

(Figure 5.4B). The presence of a putative Grb2 binding site was also observed in the extreme carboxy-terminal region of Dok-R (tyrosine residue 402), but we did not observe any association between the SH2 dornain of Grb2 and a phosphopeptide corresponding to this tyrosine residue (Figure 5.4B).

The tyrosine residues on Dok-R that are phosphorylated in response to growth factor stimulation have not been defined, so the phosphopeptide analysis represented an initial approach to define the binding sites of RasGAP and Nck on Dok-R. To determine whic h sites could become tyrosine phosphory lated upon EGF stimulation in vivo, we generated HA-tagged tyrosine to phenylalanine point mutations in full-length Dok-R at tyrosine residues 276 (Y276F), 304 (Y304F)and 351 (Y351F). We also generated compound mutations where tyrosine residues 276 and 304 were doubly mutated (DM) or IP: Dok-R Lysate EGF: = + -+ --- + RasGAP Blot: a-RasGAP

GAP-N + + + ND Nck - + + ND

Grb2 - ND ND -

Figure 5.4: Dok-R functions as a docking molecule for SH2 domain-containing proteins through phosphorylation at tyrosine residues 276, 304 and 35 1. (A) Lysates from unstimulated or EGF-stimulated COS 1 cells expressing full-length wildtype Dok- R were immunoprecipitated with anti-Dok-R antibodies. (B) The position of tyrosine residues found within YXXP motifs in the carboxy-terminal region of Dok-R are indicated (276, 304, 351) as well as one YXNX motif (402). Phosphopeptides were used for in vitro mixing expenments as descnbed in Figure 5.2C and the results are presented graphically with relative binding indicated as + or -. Interactions that were not deterrnined are denoted ND. (C)Tyrosine (Y) to phenylaianine (F)mutations were introduced in full-length HA-tagged Dok-R at the three indicated YXXP motifs and mutants were expressed in COS 1 cells and stimulated with EGF. DM (double mutant) harbors mutations at both tyrosine residues 276 and 304: TM (triple mutant) has mutations at al1 three sites. Lysates from these cells were incubated with immobilized GST-SH2 fusion proteins and bound proteins were visualized with HA im munoblotting. (D) Lysates from these cells were also irnmunopreci pitated wi th anti- Dok-R antibodies followed by anti-Nck immunoblotting to confirm the in vivo binding of Nck to tyrosine residue 35 1. Lysate Blot: a-HA

IP: HA Blot: a-pY

IP: HA

GAP4 SHZ

GAP-N+C SHZ

Nck SH2 Blot: a-HA

IP: Dok-R Blot: a-Nck al1 three tyrosine residues were mutated (TM). Al1 mutants were expressed at levels equivalent to wildtype (WT) Dok-R in COS 1 ceh and upon EGF stimulation, al1 mutants displayed similar levels of tyrosine phosphorylation and association with the EGFR

(Figure 5.4C). Because the triple mutant (TM) appeared to be phosphorylated to the same extent as wildtype Dok-R, it is likely that there are additional sites of tyrosine phosphorylation throughout the protein.

Lysates frorn these cells were then incubated with immobilized GST-SH2 domains of RasGAP and Nck and bound proteins were visualized by anti-HA immunoblotting. Both the amino-terminal (GAP-N) and the tandem SH2-SH3-SH2

(GAP-NC) dornains of RasGAP associated with al1 three single point mutants (Figure

5.4C), confirming the peptide data that showed that RasGAP can bind more than one site on Dok-R. However, the compound double mutant that Iacks both tyrosine residues 276 and 304 was no longer able to interact with the SH2 domains of RasGAP (Figure 5.4C), suggesting that RasGAP associates with Dok-R through a bisphosphorylated target motif comprising tyrosine residues 276 and 304. The SH2 domain of Nck codd associate with a phosphopeptide representing tyrosine residue 304 but mutation of this site in full-length

Dok-R had no effect on Nck binding (Figure 5.4C). In contrast, mutation of tyrosine residue 35 1 in full-length Dok-R completely abrogated binding of Nck to Dok-R (Figure

5.4C), suggesting that Nck interacted with Dok-R exclusively through this phosphorylation site. These predictions were confirmed when the compound mutations were analyzed as the Y276/304F double mutant (DM) was still able to bind Nck but the

Y276/304/35 IF triple mutant (TM) could no longer associate with Nck (Figure 5.4C). In vivo. Nck was coimmunoprecipitated with Y276F, Y304F and DM but not with Y35 IF or

TM (Figure 5.4D). In summary, these results demonstrate that upon EGF stimulation, tyrosine residues 276, 304 and 351 of Dok-R are phosphorylated and that phosphorylation at residues 276 and 304 is required for RasGAP binding while residue

35 1 is required for Nck binding. These results are in agreement with those recently reported by Lock et al. ( 1999) demonstrating that independent tyrosine residues mediate the interactions of RasGAP and Nck with Dok-R-

Recmitrnent of Dok-R to the EGFR and its association with RasGAP suggested that signaling through Dok-R might negatively regulate the Ras-MAPK pathway. To test this, we examined the effect of Dok-R expression on activation of exogenocs MAPK

-extracellular signal-regulated kinase 2 (Erk2) by immunoblotting with phospho-specific anti-Erk antibodies and in vitro phosphorylation of the substrate mye1 in basic protein

(MBP). Expression of wildtype Dok-R (WT) in COS 1 cells resulted in a marked reduction in phosphorylation of Erk2 following EGF stimulation when compared with cells expressing the vector control (Figure 5.5A). Interestingly, this suppression of Erk2 activation was not abrogated by the mutant that cannot bind RasGAP (DM) (Figure 5SA) or a mutant that cannot bind RasGAP or Nck (data not shown). As expected, the

PTB*IAPH mutant did not reduce Erk2 phosphorylation (data not shown).

The degree of Erk2-mediated phosphorylation of MBP mirrored the obsenled attenuation of Erk2 phosphorylation (Figure 5.5A). Quantitation of the levels of MBP phosphory lation revealed that in cells expressing the vector control, phosphory lation of \NT DM VECTOR EGF: O 2 5 10 02510 02510 (min)

Ppt: glutathione sepharose Blot: a-phospho-Erk?

Ppt: glutathione sepharose Blot: a-GST

Lysate Blot: a-HA

Ppt: glutathione sepharose MBP phosphorylation

Figure 5.5: Dok-R expression attenuates EGF-induced Erk2 activation independently of RasGAP. (A) GST-tagged Erk2 (O. 1 pg of plasrnid) was coexpressed in COS 1 cells with wildtype or mutant Dok-R (10 pg) and cells were left unstimulated or stirnulated with EGF for the indicated time. Lysates were precipitated (Ppt) with glutathione- sepharose and subjected to an in vitro kinase assay using myelin basic protein (MBP) as substrate. Samples were analyzed by SDS-PAGE and autoradiography or immunoblotting with anti-phospho-Erk2 antibodies. The immunoblot was reprobed with anti-GST antibodies to show approxirnately equal expression of Erk2 in al1 lanes. Non-immunoprecipitated lysates were immunoblotted with anti-HA antibodies to demonstrate equal expression of Dok-R mutants. (B) The MBP kinase activity assay was quantitated by phosphorimager analysis using the ImageQuaNT program (Molecular Dynamics): levels of MBP phosphorylation were norrnalized for relative expression of both Erk2 and Dok-R. Data shown are representative of a single experirnent which has been independently reproduced several times. EGF: (min) Erk2 was induced within 2 minutes following EGF stimulation and persisted for at least

10 minutes (Figure 5.5B). In contrast, in cells expressing wildtype Dok-R, there was 10- fold less induction of Erk2 foliowing 2 minutes of EGF stimulation and the Ievel of activation did not reach vector control levels after 10 minutes of stimulation (Figure

5.5B). Interestingly, cells expressing the RasGAP binding mutant induced Erk2 only marginal 1y more than wildtype Dok-R, suggesting that the association between Dok-R and RasGAP is not required for this attenuation of Erk activity.

As the PTB domain of Dok-R can associate in vitro with the same phosphorylated tyrosine residue on the EGFR as Shc, we wanted to confirm that attenuation of Erk activation as a result of Dok-R expression was not due to displacement of Shc from the

EGFR by Dok-R. We thus isolated Shc from unstimulated or EGF-stimulated COS 1 ceIls that were either untransfected or transfected with Dok-R. Immunoprecipitated Shc isoforms became tyrosine phosphorylated to the same extent with or without Dok-R; activated EGFR and Grb2 were readily detectable in these immunoprecipitates (Figure

5.6). Tyrosine phosphorylation of Dok-R and the associated EGFR could also be observed and Dok-R did not associate with Gr62 (Figure 5.6). These experirnents demonstrate that Shc signaling complexes are unaffected by the expression of Dok-R and the attenuation of Erk activation by Dok-R was not simply due to a loss of recruitrnent of the ShdGrb2 complex to the receptor.

Our findings are consistent with those of Nelms et al. (1 998), which demonstrate that overexpression of FRIP in 32D cells inhibits the activation of MAPK following TF: 00k-R VECTOR IP: Dok-R SHC SHC Lysate EGF: - + -+ -+ -+

. . .."-, . . ,...- -.. EGFR

Blot: a-Grb2

Figure 5.6: Shc signaling is not impaired in cells expressing Dok-R. COS 1 cells were transfected (TF)with wildtype Dok-R or the control vector, stimulated with EGF (+) or left unstimulated (-), immunoprecipitated (IP) with anti-Dok-R or anti-Shc antibodies and immunoblotted with anti-phosphotyrosine @Y) antibodies. The immunoblot was subsequently reprobed with ami-EGFR, anti-Shc and anti-Grb2 antibodies and non- immunoprecipitated lysates (Lysate) indicate the position of al1 proteins. interleukin-2 stimulation- As RasGAP negatively regulates Ras (Trahey et al., 1988), the binding of RasGAP to Dok-R is an attractive mechanism for suppression of Erk activation by Dok-R. Both Dok and Dok-R can bind RasGAP, but overexpression of

Dok has no effect on MAPK activation in response to insulin (Noguchi et al., 1999) and a mutant of Dok-R that could not bind RasGAP was unable to rescue Erk activation levels in our studies. These observations suggest that the functional differences in MAPK activation between Dok and Dok-R are probably due to unidentified binding partners that are not conserved between the two Dok family members and the physiological consequence of RasGAP binding remains to be determined. The identification of additional Dok family signaling partners will further illuminate the divergent furtctions of these docking proteins in growth factor signaling. CHAPTER 6

DISCUSSION AND CONCLUDING REMARKS. DISCUSSION AND FUTURE DIRECTIONS

Summary of the work

This study was initiated to identify further targets, substrates, and/or docking molecules that bind to the phosphorylated Tek receptor to determine exactly how Tek is eliciting its signals in endothelial cells. Following the successful isolation of a number of putative Tek binding partners, detailed analyses were performed in order to investigate the physiological relevance of these associations with Tek. Subsequent characterization of the intracellular signaling pathways utilized by these molecules was then used to establish whether these proteins could contribute to the biologica1 role of Tek in endothelial cells. Concurrent with these studies, Angl was identified as a specific activating ligand for Tek and we demonstrate that activation of Tek by Angl results in stimulation of endothelial cell survival and migration pathways. In summary, our studies have revealed that Tek can signal through a multitude of pathways and they present evidence at the moiecular and cellular level to firmly support the roles for Tek signaling in survival of the endothelium and in aspects of sprouting angiogenesis (Figure 6.1 ).

Angl Induces Endothefial Cell Survival and Migration

In the course of this study, we have identified the p85 regulatory subunit of PI 3- kinase as a Tek signaling partner and have demonstrated that activation of Tek following

Ang 1 stimulation results in tyrosine phosphorylation of p85. Tyrosine phosphorylation of p85 has been reported following the association of the tandem SH2 domains of p85 wi th numerous activated RTKs (van der Geer et al., 1994). Although the significance of this phosphorylation in PI 3-kinase activation is not clear, PI 3-kinase activity seems to Tek ENDOTHEUAL

MEMBRANE

Endothelial Endothelial

Proliferation Migration

Figure 6.1: A schematic representation surnmânzing the putative signal transduction pathways initiated upon activation of Tek by Angl. P represents a phosphorylation event. Grb 14 binding to Y 1 1 O6 is not shown. be important for Ang 1-mediated endothelial cell survival since addition of specific PI 3- kinase inhibitors abrogates the survival effect of Ang 1. Further evidence to support the roIe of Tek in mediating PI 3-kinase-dependent ce11 survival has corne from experiments by Kontos et al. (1998) where they demonstrated that activation of a chimeric Tek receptor in vivo results in increases in cellular 3' phosphoinositides and downstream activation of PKB/Akt. Moreover, Angl has been shown to protect cuttured endothelid cells from apoptosis (Kwak et al., 1999). These findings collectively suggested that

Angl could activate PI 3-kinase, resulting in induction of anti-apoptotic signal transduction pathways controlled by PKB/Akt.

In an earlier report, Witzenbichler et al. (1998) reported that a modified forrn of

Ang 1 known as Angl* does not act as a survival factor for HUVEC. An intriguing possibility that might account for this discrepancy lies in the fact that we used an essentidly unmodified forrn of Angl, with the only modification king the addition of a short carboxy-terminal epitope tag. In contrast, Ang 1 * is actually a genetically engineered chimeric ligand that consists of a short amino-terminal region of Ang2 fused to a longer carboxyl region of AngI with the inclusion of a cysteine to serine mutation at position 265 of human Angl that facilitates its purification (Koblizek et al., 1998).

AIthough this modified ligand does not exist in vivo, it appears to behave in a similar fashion to unmodified Angl in studies where both ligands have been shown to bind and activate Tek (Davis et al., 1996; Maisonpierre et al., 1997). This notwithstanding, the possible spurious outcome of using this chimeric ligand needs to be considered and the designation Ang 1* will be used herein to discuss investigations using non-native Ang 1. Over the last year, a large body of literature has become available to support the role of Ang 1 in promoting PI 3-kinase-dependent endothelial cell survival. In agreement with Our initial findings, Angl* can also stimulate tyrosine phosphorylation of p85 as well as increases in PI 3-kinase activity and inhibition of PI 3-kinase using specific inhibitors blocks the Ang 1 *-induced antiapoptotic effect (Kim et al., 2000).

Furthermore, Angl* induces PI 3-kinase-dependent activation of Akt (Fujikawa et al.,

1999: Kim et ai., 2000), it reduces apoptosis in endothelial cells (Fujikawa et al.. 1999;

Hayes et al., 1 999; Papapetropoulos et al.. 1 999) and it participates in vascular network stabilization (Papapetropoulos et al., 1999). The role of Tek in endotheliai ce11 survival complements observations in mice lacking Tek signaling pathways that display a progressive reduction in the number of Tek-expressing endothelial cells (Dumont et al.,

1994: Puri et al., 1999). Moreover, Tek has been shown to be constitutively phosphorylated in quiescent vasculature, in healing skin wounds and in a wide range of angiogenic and quiescent adult tissues, indicating that active downstream signaling is occurring in these endothelial cells (Wong et al., 1997; Yuan et al., 1999). We can therefore hypothesize that chronic activation of the PI 3-kinase signaling pathway through Tek is required for the maintenance and survival of quiescent endotheiial cells.

VEGF has also been identified as a strong survival factor for endothelial cells.

Gerber et al. (1998) have reported that activation of VEGFR-2, but not VEGFR- 1, by

VEGF can promote endothelial ceIl survival through the activation of PKB/Akt and this survival is dependent on PI 3-kinase activity. Furthermore, addition of VEGF can augment the antiapoptotic effect of Ang 1 in endothelial ceils (Kwak et al., 1999). Thus it appears as if endothelial cells have evolved to respond to two distinct survival factors,

Ang 1 and VEGF, and these survival factors share a common intracellular signaling pathway involving PI 3-kinase and PKBIAkt. It will now be interesting to determine the significance of Angl-mediated PI 3-kinase signaling in the context of a complex signaling system. Perhaps through the generation of transgenic mice that express a mutated Tek receptor that can no longer associate with p85, we can assess the contribution of Angl to PI 3-kinase-dependent endothelial ce11 survival as well as the ability of VEGF to functionally compensate for Ang 1.

Despite the evidence to suggest that Ang 1 can function as a survival factor similar to VEGF, Angl is not an endothelial mitogen like VEGF as it does not activate MAPK nor does it stimulate cellular proliferation (Davis et al., 1996; Koblizek el al., 1998;

Kwak et al., 1999; Witzenbichler et al., 1998). Interestingly, our studies as well as those of NeIms et al. (1998) have demonstrated that expression of Dok-R can attenuate MAPK activation downstream of tyrosine kinase and cytokine receptors. Moreover, DokR maps to a locus in the mouse that is rearranged in a number of lymphoproliferative disorders

(Jones and Dumont, 1998b). This region in the mouse has also been chariicterized as the licrirless (Izr) locus and it habeen suggested that reduced levels of Dok-R in hr/hr mice may contribute to the hyperproliferation of hrhr T cells (Nelms et al., 1998). Taken together, these results imply that Dok-R may function as a negative regulator of cellular pro1 i feration. Recruitmen t of Dok-R to the activated Tek receptor can therefore provide an attractive mechanism for the blockade in MAPK activation and subsequent cellular proliferation seen upon treatment of endotheiial cells with Ang 1. Our studies have also demonstrated that activation of Tek by Ang l results in stimulation of endothelial ceIl migration pathways that are partly dependent on PI 3- kinase activity. Previous reports had shown that both Ang 1 and Angl* could initiate endothelial ce11 migration and sprouting in vitro (Koblizek et al., 1998; Witzenbichler et al., 1998). More recent studies have confirmed these findings with both Angl (Kwak et al.. 1999; Teichert-Kuliszewska et al.. 2000) and Angl* (Hayes et al., 1999;

Papapetropoulos et al., 1999; Teichert-Kuliszewska et al., 2000) and Angl * can synergize with VEGF during sprouting angiogenesis in vivo (Asahara et al., 1998). This role for Angl is consistent with the findings that Tek- and Angl-nul1 mice have an angiogenic or migratory defect that is manifested as a lack of vesse1 sprouting and remodeling throughout the embryo (Dumont et al., 1994: Sato et al., 1995: Suri et al.,

1996). PI 3-kinase can mediate Ang l -dependent endothelial ceil motility (Fujikawa et czl., 1999; Jones et al., 1999). however it is interesting to note that in our studies, inhibition of PI 3-kinase activity can only partially abrogate the chernotactic effect of

Ang i on endothelial cells. This implies that additional Tek binding partners such as

Dok-R,Grb7 and Shp2 may also contribute to Ang 1-mediated endothelial ce11 migration.

Grb7 and Shp2 can prornote adhesion-dependent cell migration through associations with activated FAK (Han and Guan, 1999; Yu et al., 1998). Furthermore, a molecule related to Dok-R known as Dok has recently been shown to mediate Nck-dependent ceIl migration in response to insulin (Noguchi et al., 1999). We have identified an equivalent

Nck binding site on Dok-R, suggesting that Dok-R may also function in Nck-mediated ceIl motility downstream of the activated Tek receptor. To test this hypothesis, we are currently investigating the effect of Dok-R expression on Ang 1 -induced ce11 migration. Preliminary evidence indicates that Dok-R can in fact potentiate Angl-induced ce11 migration and this effect is abolished when the Nck binding site is deleted. Our ability to define the molecular mechanisms responsible for Ang 1 -mediated endotheiial cell motility and tubule formation further substantiates the predicted contribution of Tek signaling pathways in sprouting angiogenesis.

Tek Contains a Multîsubstrake Dociking Sire that muy be Regulared by Phosphaiases

In the majority of activated RTKs, distinct phosphotyrosine residues serve as specific binding sites for the SH2 domains of intracellular signaling proteins. We have isolated a number of SH2 domain-containing signaling molecules that can associate with

Tek and upon mapping of the potential binding sites of the SH2 dornains of these proteins on Tek, we have found that tyrosine residue 1 100 (Y1Im)can serve as a multidocking site.

Analogous multifunctional docking sites have also been described for other RTKs, including the hepatocyte growth factor receptors c-Met and STKRON as well as Ufo/Axl and the Eph receptors (Braunger et al., 1997; Hock et al.. 1998; Iwama et al.. 1996;

Ponzetto et al., 1994) suggesting that binding of multiple SH2 domain-containing proteins to a single site is a common mechanism for signal transduction by RTKs. The multisubstrate docking site on Tek is required for its association with Grb2 and Grb7 in vivo and tyrosine phosphorylation of Grb7 and p85 is abrogated when Y1lWis mutated. It should be noted that relatively high protein amounts were used in these studies, and since these levels do not reflect the physiological expression of these proteins, this may influence the data obtained. The selection of substrates for such a site is likely to be dependent on the dynamic temporal and spatial expression profiles of these proteins. Further studies examining the relative abundance of these proteins in endothelial celts will cornplement our studies of the binding affinities of these proteins for the multidocking site and will provide insight into the significance of this site in Tek signaling. Interestingly , substitution of the tandem multidocking sites in the c-Met receptor abolishes its mitogenic activity (Ponzetto et al., 1994). Given that PI 3-kinase activation has been linked to Ang 1 -mediated ce11 migration and survivai pathways, it will be interesting to determine whether mutagenesis of this site will affect Tek signaling.

Moreover, this multidocking site is also conserved in the related Tie receptor and since

Tie has also been implicated in ce11 survival (Puri et al.. 1999), future studies will undoubtedly address the possibility that p85 is recruited to this site during Tie signai transduction.

Signal transduction is a dynamic process that depends in part on the addition and removal of phosphate groups from the tyrosine residues of cell-surface receptors as well as intracellular signaling proteins. A number of tyrosine phosphatases have been identified that are expressed in endothelial ceils and we are beginning to appreciate the significant activity of these molecules in Tek signaling. Shp2 was first reported as a binding partner for Tek in 1995 (Huang et al., 1995) and we subsequently confirmed these findings using the yeast two-hybrid system. Further evidence for a role of phosphatases in Tek signaling has corne frorn technical experience where in the absence of the non-specific phosphatase inhibitor sodium orthovanadate. phosphorylation of Tek in endothelial cells is virtually undetectabie (Davis et al., 1996; Jones et al., 1999; Kontos et nl., 1998). These findings led us to believe that an active phosphatase was associated with the receptor in these cells. In support of this hypothesis, Fachinger et al. (1999) recently reported the identification of an endothelial ce Il-specific receptor-type phosphatase known as xascular endothelia1 protein tyrosine phosphatase (VE-PTP)that can specifically dephosphorylate Tek. Furtherrnore, a cytoplasmic phosphatase known as

-human cellular grotein tyosine phosphatase A (HCPTPA) has also been shown to associate with Tek and expression of HCPTPA reduces sprout formation in the rat aortic ring mode1 of angiogenesis (Huang et al., 1999).

Shp2 can become tyrosine phosphorylated in response to growth factor stimulation and this allows it to function as an adaptor molecule to recruit Grb2-SOS complexes to RTKs and positively regulate the RasMAPK pathway. However, Our studies indicate that Shp2 does not become phosphorylated when coexpressed with activated Tek thus it cannot act as an adaptor molecule in this context. Previous studies

(Huang et al., 1995) as well as Our own have demonstrated that Shp2 can associate with

Tek through phosphorylated tyrosine residue 1 1 1 1 (Y""). In order to investigate the potential role of Shp2 in Tek-mediated signaling pathways, we have generated a mutation at this site as well as other putative sites of receptor autophosphorylation and studies to characterize these mutations are currently underway. Preliminary results indicate rhat mutagenesis of both YIimand Y"W dramatically reduces receptor phosphorylation while mutation of Y'"' appears to increase the overall phosphotyrosine content of the receptor.

This striking observation implies that Y"" may function in vivo to recruit Shp2 to modulate receptor phosphorylation ievels and additional phosphatases such as HCPTPA may also contribute to this effect. These findings suggest that the differential recruitment of phosphatases to Tek may be important in attenuating the biological activity of the receptor. Intriguingly, the extracellular dornain of VE-PTP contains a series of motifs that have ken shown to participate in receptor-ligand interactions (Fachinger et al.,

1999). It will now be fascinating to see whether the angiopoietins could simultaneously bind to both Tek and VE-PTP on the surface of endothetial cells, promoting the formation of TekIVE-PTP heterodimers. In this event, VE-PTP would prevent reciprocal phosphorylation of Tek and the heterodimers would be functionally inactive.

Accordingly, if this ligand was preferentially An@ rather than Ang 1, this would provide an attractive molecular mode] for the antagonistic properties of Ang2.

The Unique Rule of Docking Molecules in Signal Transduction Paîhways

Activated ceIl surface growth factor receptors can amplify signal transduction pathways through their associations with cytoplasmic docking molecules such as IRS-1.

During the course of these studies, IRS-3, FRS2, Gab1 and Gab2 were also identified as docking molecules that associate with growth factor and cytokine receptors (Gu et al.,

1998; Kouhara et al., 1997; Lavan et al., 1997; Weidner et al., 1996; Zhao et al., 1999).

It is interesting to note that these scaffolding proteins share a similar topography consisting of distinct domains that impart a similar function in the targeting of proteins to the ce11 membrane. For instance, IRS-3 has the same general architecture as the other

IRS family proteins with an arnino-terminal PH domain and a central PTB domain although the carboxy-terminus of IRS-3 is considerably shorter (Lavan et al., 1997).

However, FRS2 contains an amino-terminal myristylation sequence and a central PTB domain and this myristylation sequence appears to fulfill a similar role as the PH domain of IRS proteins since it is required for membrane anchoring and tyrosine phosphorylation of FRS2 in FGF-stimuiated cells (Kouhara et al., 1997). Similady, Gab proteins contain an amino-terminal PH domain and a central phosphotyrosine-binding proline rich region known as the c-Met binding domain (MBD) (Holgado-Madmga et al., 1996: Weidner et al., 1996) and membrane localization of Gabl appears to be mediated by the PH dornain

(Maroun et al., 1999). The PTB domain and the MBD of these proteins further contri bute to membrane local ization for interaction with activated receptors.

The presence of both a PH domain and a PTB domain in the Dok family of proteins suggests that these proteins are likely recruited to the ce11 membrane in the vicinity of receptor binding partners during growth factor signaling. Our results demonstrate that both domains are required for an effective interaction between Dok-R and the EGFR in vivo aIthough the presence of an intact PTB domain appears to be critical in ensuring high levels of tyrosine phosphorylation of Dok-R upon EGF stimulation. It has recently been shown that the PH domain of Dok is necessary for membrane localization and tyrosine phosphorylation of Dok upon insuiin stimulation and although the PTB domain of Dok was not examined, an intact NPXY sequence in the insulin receptor was also required for this phosphorylation (Noguchi et al., 1999). In contrast to these findings, the PH domain of IRS-1 is critical for coupling to the insulin receptor (Myers et al., 1995; Voliovitch et al.. 1995; Yenush et al., 1996) and the presence of a PTB docking site in the receptor is dispensable for iRS-1 phosphorylation

(Backer et ai., 1997). The PH domain of IRS- 1 has been shown to associate reversibly with PI phosphates in the plasma membrane of unstimulated cells suggesting that localization of IRS-1 near the receptor via the PH domain may be necessary and

sufficient for phosphorylation upon insulin stimulation (Dhe-Paganon et al., 1999). The

differences between IRS and Dok proteins may reflect the ability of PH domains to

discriminate between different PI phosphates and they indica~ethat functionally related

docking molecules utilize altemate strategies for localization to ce11 surface receptors.

In addition to the membrane targeting domains of these proteins, docking

molecules also contain numerous potential tyrosine phosphorylation sites that can

mediate additional protein-protein interactions with the SH2 domains of various other

intracellular signaling proteins. Interestingly, phosphoryiation of unique families of docking molecules in response to growth factor stimulation establishes binding sites for different subsets of SH2 domain-containing signaling proteins (Figure 6.2). For instance,

Gabl appears to be the major docking molecule invoived in recruiting the p85 subunit of

PI 3-kinase to the EGFR and the nerve growth factor (NGF) receptor TrkA through its direct association with Grb2 and Grb2-Shc complexes (Holgado-Madruga et al., 1996;

Holgado-Madruga et al., 1997). Formation of such complexes allows Gabl to connect

TrkA with PI 3-kinase activation and ce11 survival (Holgado-Madruga et al., 1997) and expression of Gab 1 has also been shown to stimulate the MAPK pathway downstream of a variety of RTKs (Korhonen et al., 1999; Weidner et al., 1996). Gab2 can also bind

Grb2 and p85 and both Gab proteins can associate with Shp2 (Gu et al., 1998; Holgado-

Madruga et al., 1 996; Zhao et al., 1999), the interaction of which is essential in a parallel

Drosophila melanogaster signaling pathway (Herbst et al., 1999). Similarly, FRS2 becomes tyrosine phosphorylated in response to FGF and NGF stimulation and this Grb2 Shp2 ~85 Nck RasGAP

IRS- 1

IRS-3

Gab 1

Gab2

FRS2

Dok

Dok-R

DOKL

Figure 6.2: Summary of the SH2 domain-containing signaling molecules that associate with various phosphorylated docking molecules. Each + represents an interaction that has been shown to occur either in vitro or in vivo between the isolated SH2 domains and the indicated docking moIecules. Each - represents an interaction that does not occur while ND indicates that the interaction has not been tested, detemined or reported. Adapted from references in the text. facilitates its association with Grb2, both directly and indirectly through Shp2 (Hadari et

al., 1998; Kouhara et al., 1997; Meakin et al., 1999). Since both the FGF and NGF

receptors lack the consensus YXNX motif required for Grb2 binding, FRS2 plays a

principal role in mediating signals from activated receptors to the RasMAPK signaling

cascade (Hadari et al., 1998; Kouhara et al., 1997). Phosphorylated IRS proteins have

also been shown to associate with a similar repertoire of signaling proteins, including

Grb2, Shp2 and p85 as weli as the Nck adaptor protein (Ross et al., 1998; Xu et al,. 1999:

Yenush and White, 1997).

In marked contrast to many other families of docking proteins, Dok proteins do

not appear to associate with Grb2 and instead, phosphorylation of both Dok and Dok-R

results in complex formation with Nck (Holland et al.. 1997; Jones and Dumont, 1998a;

Noguchi et al., 1999; Tang et al., 1997) and RasGAP (Carpino et al., 1997: Di Cristofano et czl., 1998; Holland et al., 1997; Jones and Dumont. 1998a; Nelms et al., 1998: Noguchi et al.. 1999; Yamanashi and Baltimore, 1997). Many RTKs do not contain the consensus

YDEP or extended YXXPXD motifs required for optimal binding of Nck and RasGAP

respectively (Hu and Settleman, 1997; Songyang et al.. 1993) and recruitment of a

docking molecule such as Dok or Dok-R is likely required to effectively transduce

signals through these proteins. Binding of Nck to Dok proteins seems to be required for

ceIl motility (Noguchi et al., 1999 and our unpublished work) while the role of RasGAP

binding to Dok family proteins is less clear. RasGAP can decrease the activity of Ras. - thus downregulating signaling through the Ras-MAPK pathway. Despite the ability of

Dok to bind RasGAP, expression of Dok has previously ken reported to have no effect on MAPK activation (Cong et al., 1999; Noguchi et al., 1999). However, primary B cells isolated from mice lacking Dok have recently beeri shown to have higher levels of

MAPK activation than normal B cells following antigenic stimulation of the B ce11 receptor (Yamanashi et al., 2000). These experiments indicate that Dok may actually function as a negative regulator of MAPK, Dok-R can also attenuate MAPK activation

(Nelms er al., 1998) and this effect appears to be surprisingly independent of RasGAP binding in Our studies. RasGAP can also mediate changes in cell shape and adhesion

(McGlade et al., 1993) and our findings suggest that the fraction of Dok- and Dok-R- bound RasGAP could potentially function to mediate cytoskeletal rearrangements in collaboration with Nck. It is interesting to note that a newly isolated member of the Dok farnily referred to as DOKL (for p62""'-like protein) does not associate with RasGAP or

Nck although DOKL functions to downregulate MAPK activation in response to certain stimuli (Cong et al., 1999). Collectively these experiments implicate the Dok proteins in negative regulation of ce11 signaling through a RasGAP-independent mechanism of Ras-

MAPK suppression. The Dok proteins may also have divergent functions in particular ce1 l lines since the remaining putative tyrosine phosphorylation sites are not well conserved and thus are likely to interact with unique SW2 domain-containing signaling molecules.

Taken together, these studies imply that scaffolding substrates are a conserved mechanism for signal amplification downstream of a number of unrelated RTKs and that despite overlapping binding profiles between certain docking molecules, each unique protein complex likely functions to mediate a distinct biological effect. CONCLUDING REMARKS

A common property amongst angiogenic ligands such as VEGF and the angiopoietins is that they cm elicit multiple responses depending upon the context of their expression and the presence of other growth factors. Study of the signal transduction pathways mediated by these growth factors provides insight into the molecular and cellular mechanisms that regulate vesse1 assembly. Key components of signal transduction cascades cmthen be used as potential targets in angiogenic therapies.

Altered angiogenesis is manifested by a number of important diseases and the ability to manipulate specific pathways may have implications in the therapeutic control of vascular growth. For instance, the ability to locally inhibit signaling pathways that play a role in the survival of the endothelium could potentially disrupt the pathological angiogenesis seen in cancer and diabetic retinopathy. Moreover, enhancement of signaling pathways that promote sprouting angiogenesis would serve as an approach to the vascular insufficiency seen in ischemic and injured tissues. Angl/Tek signaling pathways have been shown to function in the maintenance and survival of endothelial cells as well as in the expansion of the vasculature. The ability to tightly regulate these pathways will undoubtedly continue to be of great clinical importance in the future management of angiogenic diseases. BIBLIOGRAPHY Alessi, D. R., Andjelkovic, M., Caudwell, B., Cron, P., Motrice, N., Cohen, P., and Hemrnings, B. A. (1996). ~Mechanismof activation of protein kinase B by insulin and IGF- 1. EMBO '15,654 1-6551.

Alessi, D. R., James, S. R., Downes, C. P., Holmes, A. B., Gaffney, P. R., Reese, C. B., and Cohen, P. (1997). Characterization of a 3-phosphoinositide-dependent protein kinase which phosphorylates and activates protein kinase Baipha. Curr. Biol. 7,26 1-269.

Alon, T., Hemo, I., Itin, A., Pe'er, J., Stone, J., and Keshet, E. (1995). Vascular endothelid growth factor acts as a survival factor for newly fonned retinal vessels and has implications for retinopathy of prematurity. Nat. Med. 1, 1024-1028.

Anderson, D., Koch, C. A., Grey, L., Ellis, C., Moran, M. F., and Pawson, T. (1990). Binding of SH2 domains of phospholipase C gamma 1, GAP, and Src to activated growth factor receptors. Science. 250,979-982.

Andjelkovic, M., Alessi, D. R., Meier, R., Fernandez, A., Lamb, N. J., Frech, M., Cron, P., Cohen, P., Lucocq, J. M., and Hemmings, B. A. (1997). Role of translocation in the activation and function of protein kinase B. J. Biol. Chem. 272,3 15 15-3 1524.

Antonelli-Orlidge, A., Saunders, K. B., Smith, S. R., and D'Amore, P. A. (1989). An activated fonn of transforming growth factor beta is produced by cocultures of endothelial ceIIs and pericytes. Proc. Natl. Acad. Sci. U.S. A. 86,4544-4548.

Asahara, T., Murohara, T., Sullivan, A., Silver, M., van der Zee, R., Li, T., Witzenbichler, B., Schatteman, G., and Isner, J. M. (1997). Isolation of putative progenitor endothelial cells for angiogenesis. Science. 275,964-967.

Asahara, T.. Chen, D., Takahashi, T., Fujikawa, K., Keamey, M., Magner, M., Yancopoulos, G. D: and Isner, J. M. (1998). Tie2 receptor ligands, angiopoietin-1 and angiopoietin-2, modulate VEGF- induced postnatal neovascularization. Circ. Res. 83, 233 -240.

Auerbach, R., Alby, L., Momssey, L. W., Tu, M., and Joseph, J. (1985). Expression of organ-specific antigens on capillary endothelial cells. Microvasc. Res. 29,40 1-4 1 1.

Ausprunk, D. H. and Folkman, J. ( 1977). Migration and proliferation of endothelial cells in preformed and newly forrned blood vessels during tumor angiogenesis. Microvasc. Res. 14,53-65.

Backer, J. M., Wjasow, C., and Zhang, Y. (1997). In vitro binding and phosphorylation of insulin receptor substrate 1 by the insulin receptor. Role of interactions mediated by the phosphotyrosine-binding domain and the pleckstrin-homology domain. Eur. J. Biochem. 245,9 1-96. Beck, L. and D'Amore P. A. (1997). Vascular development: cellular and molecular recognition. FASEB J. 11, 365-373.

Benjamin, L. E., Hemo, I., and Keshet, E. (1998). A plasticity window for blood vessel remodelling is defined by pericyte coverage of the preforrned endothelia! network and is regulated by PDGF- B and VEGF. Development. 125, 1591 - 1598.

Bernard, O., Barin, C., Charrin, C., Mathieu-Mahul, D., and Berger, R. (1993). Characterization of translocation t( 1 ;14)(p32;q 1 1) in a T and in a B acute leukemia. Lertkenzia. 7, 1509- 15 13.

Birge, R. B., Knudsen, B. S., Besser, D., and Hanafusa, H. (1996)- SH2 and SH3- containing adaptor proteins: redundant or independent mediators of intracellular signal transduction. Genes, Cells. 1, 595-6 13.

Blaikie, P., Immanuel, D., Wu, J., Li, N., Yajnik, V., and Margolis, B. (1994). A region in Shc distinct from the SH2 domain can bind tyrosine- phosphorylated growth factor receptors. J. Biol. Chem. 269, 3203 1-32034.

Borg, J. P., Ooi, J., Levy, E., and MargoIis, B. (1996). The phosphotyrosine interaction domains of XI 1 and FE65 bind to distinct sites on the YENPTY motif of amyloid precursor protein. Mol. Cell BioZ. 16,6229-6241.

Braunger, J., Schleithoff, L., Schulz, A. S., Kessler. H., Lamrners, R., Ullrich, A., Bartram, C. R., and Janssen, J. W. (1997). Intracellular signaling of the Ufo/Axl receptor tyrosine kinase is mediated mainly by a multi-substrate docking-site. Oncogene. 14, 26 19-263 1.

Calvert, J. T., Riney, T. J., Kontos, C. D., Cha, E- H., Prïeto, V. G-,Shea, C. R., Berg, J. N., Nevin, N. C., Simpson, S. A., Pasyk, K. A., Speer, M. C., Peters, K. G., and Marchuk, D. A. ( 1999). Allelic and locus heterogeneity in inherited venous malformations. Hum. Mol. Genet. 8, 1279- 1289.

Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt. C.. Declercq, C., Pawling, J., Moons, L., Collen, D., Risau, W., and Nagy, A. ( 1996). Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature. 380,435- 439.

Carmeliet, P., Moons, L., and Collen, D. (1998). Mouse models of angiogenesis, arterial stenosis, atherosclerosis and hemostasis. Cardiovasc. Res. 39,8-33.

Canneliet, P., Ng, Y. S., Nuyens, D., Theilmeier, G., Brusselmans, K., Cornelissen, I., Ehler, E., Kakkar, V. V., Stalmans, I., Mattot, V., Pemard, J. C., Dewerchin, M., Flameng, W., Nagy, A., Lupu, F., Moons, L., Collen, D., D'Amore, P. A., and Shima. D. T. ( 1999). Impaired myocardial angiogenesis and ischemic cardiornyopathy in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Nat. Meù. 5,495-502.

Carpino, N., Wisniewski, D., Strife, A., Marshaic, D., Kobayashi, R., Stillman, B., and Clarkson, B. (1997). p62(dok): a constitutively tyrosine-phosphorylated, GAP-associated protein in chronic myelogenous leukemia progenitor cells. Ce11 88, 197-204.

Charest, A., Wagner, J., Jacob, S., McGlade, C. J., and Tremblay, M. L. (1996). Phosphotyrosine-independent binding of SHC to the NPLH sequence of murine protein- tyrosine phosphatase-PEST. Evidence for extended phosphotyrosine bindi iig/phosphotyrosine interaction domain recognition speci ficity . J. Biol. Chem. 271, 8424-8429.

Cheung, A. H., Stewart, R. J., and Marsden, P. A. (1998). Endothelial TieYïek ligands angiopoietin- 1 (ANGPT 1) and angiopoietin-2 (ANGPTS): regional localization of the human genes to 8q22.3-q23 and 8p23. Genomics. 48,389-39 1.

Chien, C. T., Bartel, P. L., Stemglanz, R., and Fields, S. (1991). The two-hybrid systern: a method to identify and clone genes for proteins that interact with a protein of interest. Proc. Narl. Acad. Sci. U. S. A. 88,9578-9582.

Chien, C. T., Wang, S., Rothenberg, M., Jan, L. Y., and Jan. Y. N. (1998). Numb- associated kinase interacts with the phosphotyrosine binding dornain of Numb and antagonizes the function of Numb in vivo. Mol. Ce11 Biol. 18,598-607.

Claesson-Welsh, L. (1999). VascuIar growth factors and angiogenesis. Curr. Top. Microbiol. Immunol. 237, 1 - 1 87.

Cofîin, J. D. and Poole, T. J. ( 1988). Embryonic vascular development: immunohistochemicai identification of the origin and subsequent morphogenesis of the major vesse1 primordia in quai1 embryos. Development. 102, 735-748.

Cong, F., Yuan, B., and Goff, S. P. (1999). Characterization of a novel member of the DOK family that binds and modulates Ab1 signaling. Mol. Cell Biol. 19, 83 14-8325.

Cooper, J. A. and Kashishian, A. (1993). In vivo binding properties of SH2 domains from GTPrise-activating protein and phosphatidylinositol 3-kinase. Mol. Cell Biol. 13, 1737- 1745.

Cunningham, S. A., Waxham, M. N., Arrate, P. M., and Brock, T. A. (1995). Interaction of the Fit-I tyrosine kinase receptor with the p85 subunit of phosphatidylinositol 3- kinase. Mapping of a novel site involved in binding. J. Biol. Chern. 270,20254-20257. D'Amore, P. A. (1992). Capillary growth: a two-cell system. Sernin. Cancer Biol. 3,49- 56.

Datta, S. R., Dudek, H*,Tao, X., Masters, S., Fu, H., Gotoh, Y., and Greenberg, M. E. ( 1997). Akt phosphorylation of BAD couples survival signals to the cd-intrinsic death machinery. Cell. 91,231-24 1.

Davis, S., Aldrich, T. H,,Jones, P. F., Acheson, A., Compton, D. L., Jain, V., Ryan, T. E., Bruno, J., Radziejewski, C., Maisonpierre, P. C., and Yancopoulos, G. D. (1996). Isolation of angiopoietin- 1, a ligand for the TE2 receptor, by secretion-trap expression cloning. Ce11 87, 1 16 1- 1169.

del Peso, L., Gonzaiez-Garcia, M., Page, C.. Herrera. R., and Nunez, G. (1997). Interleukin-3-induced phosphoryiation of BAD through the protein kinase Akt. Science. 278. 687-689.

Dhe-Paganon, S., Ottinger, E. A., Nolte, R. T., Eck, M. J., and Shoelson, S. E. (1999). Crystal structure of the pleckstrin homology-phosphotyrosine binding (PH-PTB) targeting region of insulin receptor substrate 1. Proc. Nd.Acad. Sci. CI. S. A. 96, 8378- 8383.

Dho, S. E., Jacob, S., Wolting, C. D., French, M. B., Rohrschneider, L. R., and McGlade, C. J. ( 1 998). The rnarnmalian numb phosphotyrosine-binding domain. Characterization of binding specificity and identification of a novel PDZ domain-containing numb binding protein, LNX.J. Biol. Chern. 273,9179-9 187.

Di Cristofano, A., Carpino, N., Dunant, N., Friedland, G., Kobayashi, R., Strife, A., Wisniewski, D., Clarkson, B., Pandolfi, P. P., and Resh, M. D. (1998). Molecular cloning and characterization of p56dok-2 defines a new farnily of RasGAP-binding proteins. J. Biol. Clrern. 273.4827-4830.

Downward, J,, Parker, P., and Waterfield, M. D. (1984). Autophosphorylation sites on the epidermal growth factor receptor. Nature. 311,483-485.

Downward, J. ( 1998). Mechanisms and consequences of activation of protein kinase B/Akt. Cccrr. Opin. Ce11 Biol. 10, 262-267.

Dubois, N. A., Kolpack. L. C., Wang, R., Azizkhan, R. G., and Bautch, V. L. ( 199 1). Isolation and characterization of an established endothelial ce1 l line from transgenic mouse hemangiomas. Exp. Cell Res. 196,302-3 13.

Dudek, H., Datta, S. R-,Franke, T. F., Birnbaum, M. J., Yao, R., Cooper, G. M.. Segal, R. A., Kaplan, D. R., and Greenberg, M. E. (1997). Regdation of neuronal survival by the serine-threonine protein kinase Akt. Science. 275,66 1-665. Dumont, D. J., Yamaguchi, T. P., Codon, R. A., Rossant, J., and Breitman, M. L. (1992). tek, a novel tyrosine kinase gene located on mouse chromosome 4, is expressed in endothelial cells and their presumptive precursors. Oncogene. 7, 147 1-1480.

Dumont, D. J., Gradwohl, G. J., Fong, G-H., Auerbach, R., and Breitman, M. L. (1993). The endothelial-specific receptor tyrosine kinase, tek, is a member of a new subfamily of receptors. Oncogene. 8, 1293- 130 1.

Dumont, D. J., Gradwohl, G., Fong, G. H., Puri, M. C., Gertsenstein, M., Auerbach, A.. and Breitman, M. L. ( 1994). Dominant-negative and targeted nul1 mutations in the endotheliai receptor tyrosine kinase, tek, reveal a critical role in vasculogenesis of the embryo. Genes. Dev. 8, 1897- 1909.

Dumont, D. J., Fong, G. Fi., Puri, M. C., Gradwohl, G., Alitalo, K., and Breitman, M. L. ( 1995). Vascularization of the mouse embryo: a study of flk- 1, tek, tie. and vascular endotheliai growth factor expression during development. Dev. Dyn. 203,80-92.

Dumont, D. J., Jussila, L. -F., Seet, T., Mustonen, T., and Alitalo, K-(1998). Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science. 282, 946-949.

Eck. M. J., Shoelson, S. E., and Harrison, S. C. (1993). Recognition of a high-affinity phosphotyrosyl peptide by the Src homology-2 domain of p56lck. Nature. 362, 87-9 1.

Eck, M. J., Dhe-Paganon, S., Tmb, T., Nolte, R. T., and Shoelson, S. E. (1996). Structure of the IRS- 1 PTB domain bound to the juxtamembrane region of the insulin receptor. Cell85,695-705.

Ellis. C., Moran. M., McCorrnick, F., and Pawson, T. (1990). Phosphorylation of GAP and GAP-assoçiated proteins by transforming and mitogenic tyrosine kinases. Nature. 343,377-38 1.

Engerrnan, R. L., Pfaffenbach, D., and Davis, M. D. (1967). Ce11 turnover of capillaries. Lab. Invest. 17,738-743.

Fachinger, G., Deutsch, U., and Risau, W. (1999). Functional interaction of vascular endothelial-protein-tyrosine phosphatase with the tie-2. Oncogene. 18, 5948-5953.

Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O'Shea, K. S., Powell- Braxton, L., Hillan, K. J., and Moore, M. W. (1996). Heterozygous ernbryonic lethality induced by targeted inactivation of the VEGF gene. Nature. 380,439-442.

Ferrara, N. and Alitalo, K. (1999). Clinical applications of angiogenic growth factors and their inhibitors. Nat. Med. 5, 1359-1364. Ferrara, N. ( 1999). Vascular endothelial growth factor: molecular and bioiogical aspects. Cu rr. Top. Microbiol. Immunof. 237, 1-30.

Fiddes, R. J., Campbell, D. H., Janes, P. W., Sivertsen, S. P., Sasaki, H., Wailasch, C., and Daly, R. J. ( 1998). Analysis of Grb7 recniitment by heregulin-activated erbB receptors reveals a novel target selectivity for erbB3. J. Biol. Chem. 273, 7717-7724.

FieIds, S. and Song, 0. ( 1989). A novel genetic system to detect protein-protein interactions. Nature. 340.245-246.

Folkman, J. (197 1). Tumor angiogenesis: therapeutic implications. N. Engl J. Med. 285? 1182-1 186.

FoIkman, J. ( 1995). Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat. Med. 1, 27-3 1.

Fong, G. H., Rossant, J., Gertsenstein. M., and Breitman. M- L. (1995). Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nanrre. 376. 66-70.

Franke, T. F., Kaplan, D. R., Cantley, L. C., and Toker, A. ( 1997). Direct regulation of the Akt proto-oncogene product by phosphatidy linosi tol-3 +bisphosphate. Science. 275, 665-668.

Frech, M., Andjelkovic, M., Ingley, E., Reddy, K. K., Falck, J. R., and Hemmings, B. A. ( 1997). High affrnity binding of inositol phosphates and phosphoinositides to the pleckstrin hornology domain of RAC/protein kinase B and their influence on kinase activity . J. Biol. Chem. 272, 8474-848 1.

Fujikawa, K., de Aos Scherpenseel, I., Jain, S. K.. Presman, E., and Varticovski, L. ( 1999). Role of PI 3-kinase in angiopoietin-1 -mediated migration and attachment- dependent survival of endothelial cells. Exp. Ce11 Res. 253,663-672.

Garrity, P. A., Rao, Y., Salecker, I., McGlade. J., Pawson, T., and Zipursky, S. L. (1996). Drosophila photoreceptor axon guidance and targeting requires the dreadlocks SH2/SH3 adapter protein. Cell85,639-650.

Gerber, H. P., McMurtrey, A., Kowalski, J., Yan, M., Keyt, B. A., Dixit, V., and Ferrara, N. ( 1998). Vascular endothelial growth factor regulates endothelial ce11 suwival through the phosphatidylinositd 3'-kinase1Akt signal transduction pathway. Requirernent for Flk- 1 IKDR activation. J. Biol. Chem. 273, 30336-30343.

Gietz, D., St-Jean, A., Woods, R. A., and Schiestl, R. H. ( 1992). improved method for high efficiency transformation of intact yeast cells. Nucleic Acids Res. 20, 1425. Gluzrnan, Y. ( 198 1 ). SV40-transformed simian cells support the replication of early SV40 mutants. Ce11 23, 175- 182.

Gotoh, T., Hattori, S., Nakamura, S., Kitayama, H., Noda, M., Takai, Y., Kaibuchi, K.. Matsui, H., Hatase, O., and Talcahashi, H. (1995). Identification of Rapl as a target for the Crk SH3 domain-binding guanine nucleotide-releasing factor C3G. Mol. Cell Biot. 15,6746-6753-

Gu, H., Pratt, J. C., Burakoff, S. J., and Neel, B. G. (1998)- Cloning of p97/Gab2, the major SHP2-binding protein in hematopoietic cells, reveals a novel pathway for cytokine- induced gene activation. Mol. Cell2, 729-740.

Gustafson, T. A., He, W., Craparo, A., Schaub, C. D., and O'Neill, T. J. (1995). Phosphotyrosine-dependent interaction of SHC and insulin receptor substrate 1 with the NPEY motif of the insulin receptor via a novel non- SH2 domain. Mol. Cell Biol. 15. 2500-2508.

Hadari, Y. R., Kouhara, H., Lax. I., and Schlessinger, J. (1998). Binding of Shp2 tyrosine phosphatase to FRS2 is essential for fibroblast growth factor-induced PC12 ceIl differentiation. Mol. Ce11 Biol. 18, 3966-3973.

Han, D. C. and Guan, J. L. (1999). Association of focal adhesion kinase with Grb7 and its role in ce11 migration. J. Biol. Chem. 274,24425-24430.

Hanahan, D. ( 1997). Signal ing vascular morphogenesis and maintenance. Science. 277, 48-50.

Hanks, S. K. and Polte, T. R. (1997). Signaling through focal adhesion kinase. Bioassuys. 19, 137-145.

Hatzopoulos, A. K., Fotkman, J., Vasile, E., Eiselen, G. K., and Rosenberg, R. D. (1998). Isolation and characterization of endothelial progenitor cells from mouse embryos. Developrnent. 125, 1457- 1468.

Hayes, A. J., Huang, W. Q., Mallah, J., Yang, D., Lippman, M. E., and Li, L. Y. (1999). Angiopoietin- 1 and its receptor tie-2 participate in the regulation of capillary-like tubule formation and survival of endothelia1 cells. Microvasc. Res. 58, 224-237.

He, W., Rose, D. W., Olefsky, J. M., and Gustafson, T. A. (1998). GrblO interacts differentially with the insulin receptor, insulin-like growth factor 1 receptor, and epidermal growth factor receptor via the Grb 10 Src homology 2 (SH2) domain and a second novel domain located between the pleckstrin homology and SH2 domains. J. Biol. Cllem. 273,6860-6867. Henkemeyer, M., Rossi, D. J., Holmyard, D. P., Puri, M. C.,Mbarnalu, G., Harpal, K., Shih. T. S., Jacks, T., and Pawson, T. (1995). Vascular system defects and neuronal apoptosis in mice lacking ras GTPase-activating protein. Nature. 377,695-70 1.

Herbst, R., Zhang, X., Qin, J., and Simon, M. A. (1999). Recruitment of the protein tyrosine phosphatase CSW by DOS is an essential step during signaling by the sevenless receptor tyrosine kinase. EMBO J. 18,6950-6961,

Hiratsuka, S., Minowa, O., Kuno, J., Noda, T., and Shibuya, M. (1998). Flt-1 lacking the tyrosine kinase domain is sufficient for normal deve!opment and angiogenesis in mice. Proc. Natl. Acad, Sci. U.S. A. 95,9349-9354.

Hock, B., Bohme, B., Karn, T., Feller, S., Rubsamen-Waigmann, H.,and Strebhardt, K. ( 1998). Tyrosine-6 14, the major aufophosphorylation si te of the receptor tyrosine kinase HEU,functions as multi-docking site for SH2-domain mediated interactions. Oncogene. 17,255-260.

Hof, P., Pluskey, S., Dhe-Paganon, S., Eck, M. J., and Shoelson, S. E. (1998). Crystal structure of the tyrosine phosphatase SHP-2. Ce11 92,441-450.

Holash, J., Maisonpierre, P. C., Compton, D., Boland, P., Alexander, C. R., Zagzag, D., Yancopoulos, G. D., and Wiegand, S. J. (1999). Vessel cooption, regression. and growth in tumors mediated by angiopoietins and VEGF. Science. 284, 1994- 1998.

Holgado-Madmga, M., Emlec, D. R., Moscatello, D. K., Godwin, A. K., and Wong, A. J. ( 1996). A Grb2-associated docking protein in EGF- and insulin-receptor signalling. Nrrtrrre. 379,560-564.

Holgado-Madmga, M., Moscatello, D. K., Emlet, D. R., Dieterich, R., and Wong, A. J. ( 1997). Grb2-associated binder- 1 mediates phosphatidylinositol 3-kinase activation and the promotion of cell survival by nerve growth factor. Proc. Natl. Acad. Sci. U. S. A. 94, 1 24 1 9- 12424.

Holland, S. J., Gale, N. W., Gish, G. D., Roth, R. A., Songyang, Z., Cantley. L. C., Hen kemeyer, M., Yancopoulos, G. D., and Pawson, T. ( 1997). Juxtamembrane tyrosine residues couple the Fph family receptor EphB2Nuk to specific SH2 domain proteins in neuronal cells. EMBO J. 16,3877-3888.

Horita, K., Yagi, T., Kohmura, N-,Tomooka, Y., ikawa, Y., and Aizawa, S. (1992). A novel tyrosine kinase, hyk, expressed in murine embryonic stem ceIls. Biochem. Biopltys. Res. Cornrnrtn. 189, 1747- 1753.

Hu, K. Q. and Settleman, J. (1997). Tandem SH2 binding sites mediate the RasGAP- RhoGAP interaction: a conformational mechanism for SH3 domain regulation. EMBO J. 16,473-483. Huang, L., Turck, C. W., Rao, P., and Peters, K. G. (1995). GRB2 and SH-PTP2: potentially important endotheliai signaling molecules downstrearn of the TEKîTIE2 receptor tyrosine kinase. Oncogene. 11,2097-2 103.

Huang, L., Sankar, S., Lin, C., Kontos, C. D., Schroff, A. D., Cha, E. H., Feng, S. M., Li, S. F., Yu, Z., Van Etten, R. L., Blanar, M. A., and Peters, K. G. (1999). HCPTPA, a protein tyrosine phosphatase that regulates vascular endothelial growth factor receptor- mediated signai transduction and biologicai activity. J. Biol. Chem. 274,38 183-38 188.

Iwama, A.. Hamaguchi, I., Hashiyama, M., Murayama, Y., Yasunaga, K., and Suda, T. ( 1993). Molecular cloning and characterization of mouse TE and TEK receptor tyrosine kinase genes and their expression in hematopoîetic stem cells. Biochern. Biophys. Res. Conirnun. 195, 30 1-309.

Iwarna, A., Yamaguchi, N., and Suda, T. (1996). STWRON receptor tyrosine kinase mediates both apoptotic and growth signals via the multifunctional docking site conserved among the HGF receptor family. EMBO J. 15,5866-5875.

Janes, P. W., Lackrnann, M., Church, W. B., Sanderson, G. M., Sutherland, R. L., and Daly. R. J. (1997). Structural determinam of the interaction between the erbB2 receptor and the Src homology 2 domain of Grb7. J. Biol. Chem. 272,8490-8497.

Jones, N. and Dumont, D. J. (1998a). The Ternie2 receptor signais through a novel Dok- related docking protein, Dok-R.Oncogene. 17, 1097- 1 108.

Jones, N. and Dumont, D. J. (1998b). Assignment of the Dokr gene to mouse chromosome 14D2-D3 by fluorescence in situ hybridization. Genornics. 53,413-4 14.

Jones. N. and Dumont, D. J. (1999). Recruitment of Dok-R to the EGF receptor through its PTB domain is required for attenuation of Erk MAP kinase activation. Crtrr. Biol. 9, 1057- 1060.

Jones, N., Master, Z., Jones, J., Bouchard, D., Gunji, Y., Sasaki, H., Daly, R., Alitalo, K., and Dumont, D. J. (1999). Identification of TekITie2 binding partners. Binding to a rnultifunctional docking site mediates ce11 survival and migration. J. Biol. Clzent. 274, 30896-30905.

Joukov, V., Pajusola, K., Kaipainen, A., Chilov, D., Lahtinen, I., Kukk, E., Sakseh, O., Kalkkinen, N., and Alitalo, K. (1996). A novel vascular endotheliai growth factor, VEGF-C, is a ligand for the Flt4 (VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases. EMBO J. 15,290-298.

Kaipainen, A., Korhonen, J., Mustonen, T., van Hinsbergh, V. W., Fang, G. H., Dumont, D., Breitman, M., and Alitalo, K. (1995). Expression of the fms-like tyrosine kinase 4 gene becomes restncted to lymphatic endothelium during development. Proc. NatL Acad. Sci. U.S. A. 92,3566-3570.

Kasus-Jacobi, A., Perdereau, D., Auzan, C., Clauser, E., Van Obberghen, E., Mauvais- Jarvis, F., Girard, J., and Bumol, A. F. (1998). Identification of the rat adapter Grb14 as an inhibitor of insulin actions. J. Biol. Chern. 273,26026-26035.

Kavanaugh, W. M. and Williams, L. T. (1994). An alternative to SH2 domains for binding tyrosine-phosphory lated proteins. Science. 266, 1862- 1865.

Kavanaugh, W. M., Turck, C. W., and Williams, L. T. (1995). PTB domain binding to signal hgproteins through a sequence motif contain hg phosphotyrosine. Science. 268, 1177-1 179.

Keegan, K. and Cooper, J. A. (1996). Use of the two hybrid system to detect the association of the protein- tyrosine-phosphatase, SHPTP2, with another SH2-containing protein, Grb7. Oncogene. 12, 1537- 1544.

Keilhack, H., Tenev, T., Nyakatura, E., Godovac-Zimmermann, J., Nielsen, L., Seedorf, K.. and Bohmer, F. D. (1998). Phosphotyrosine 1 173 mediates binding of the protein- tyrosine phosphatase SHP- I to the epidermal growth factor receptor and attenuation of receptor signaling. J. Biol. Chern. 273, 24839-24846.

Kim, I., Kwak,H. J., Ahn, J. E.,So, J. N.,Liu, M.,Koh, K. N.,andKoh,G. Y.(1999a). Molecular cloning and characterization of a novel angiopoietin family protein, angiopoietin-3. FEBS Lm. 443, 353-356.

Kim, I., Moon, S. O., Koh. K. N., Kim, H., Uhm, C. S., Kwak, H. J., Kim, N. G., and Koh, G. Y. ( 1999b). Molecuiar cloning, expression, and characterization of angiopoietin- related protein. angiopoietin-related protein induces endothelial ce11 sprouting. J. Biol. Chm. 274,26523-26528.

Kim, 1.. Kim, H. G., So, J. -Ne,Kim, J. H., Kwak, H. J., and Koh, G. Y. (2000). Angiopoietin-1 regulates endothelial ceIl survival through the phosphatidylinositol 3'- kinase/Akt signal transduction pathway. Circ. Res. 86, 24-29.

Kitsukawa, T., Shimono, A., Kawakami, A., Kondoh, H., and Fujisawa, H. (1995). Overexpression of a membrane protein, neuropilin, in chimeric mice causes anomalies in the cardiovascular system, nervous system and limbs. Developrnent. 121,4309-43t 8.

Kitsukawa, T., Shimizu, M., Sanbo, M., Hirata, TT.,Taniguchi, M., Bekku, Y.. Yagi, T., and Fujisawa, H. (1997). Neuropilin-semaphorin IWD-mediated chemorepuIsive signais play a crucial role in peripheral nerve projection in mice. Neuron. 19,995-1005. Klippel, A., Escobedo, J. A., Fantl, W. J., and Williams, L. T. (1992). The C-terminal SH2 domain of p85 accounts for the high affinity and specificity of the binding of phosphatidylinositol 3-kinase to phosphorylated platelet-derived growth factor beta receptor. Mol. Cell Biol. 12, 145 1- 1459.

Klippel, A., Kavanaugh, W. M., Pot, D., and Williams, L. T. (1997). A specific product of phosphatidylinositol 3-kinase directly activates the protein kinase Akt through its pleckstrin homology dornain. Mol. Cell Biol. 17, 338-344.

Koblizek, T. I., Runting, A. S., Stacker, S. A., Wilks, A. F., Risau, W., and Deutsch, U. (1997). Tie2 receptor expression and phosphorylation in cultured cells and mouse tissues. Ertr. J. Biochem, 244, 774-779.

Koblizek, T-I., Weiss, C., Yancopoulos, G. D., Deutsch, U., and Risau, W. (1998). Angiopoietin- l induces sprouting angiogenesis in vitro. Curr. Biol. 8,529-532.

Koch, C. A., Anderson, D., Moran, M. F., Ellis, C.. and Pawson, T. (1991). SH2 and SH3 domains: elements that control interactions of cytoplasmic signaling proteins. Science. 252, 668-674.

Kontos, C. D., Stauffer, T. P., Yang, W. P., York, J. D., Huang, L., Blanar. M. A.. Meyer, T., and Peters, K. G. (1998). Tyrosine 1 101 of Tie2 is the major site of association of p85 and is required for activation of phosphatidylinositol 3-kinase and Akt. Mol. CdBiol. 18.4131-4140.

Korhonen, J. M., Said, F. A., Wong, A. J., and Kaplan, D. R. (1999). Gab 1 mediates neurite outgrowth. DNA synthesis, and survival in PC 12 cells. J. Biol. Cltem. 274, 37307- 373 14.

Korpelainen, E. I., Karkkainen, M., Gunji, Y., Vikkula, M., and Alitalo, K. (1999). Endothelial receptor tyrosine kinases activate the STAT signaling pathway: mutant Tie-2 causing venous malformations signals a distinct STAT activation response. Oncogene. 18. 1-8.

Kouhara, H., Hadari, Y. R., Spivak-Kroizman, T., Schilling, J., Bar-Sagi, D., Lax, I., and Schlessinger, J. ( 1997). A lipid-anchored Grb2-binding protein that links FGF-receptor activation to the RastMAPK signaling pathway . Cell89, 693-702.

Kozak, M. (1986). Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell44, 283-292.

Kundra.V.,Escobedo,J.A., Kazlauskas, A., Kim, H. K.,Rhee, S. G., Williams, L.T., and Zetter, B. R. (1994). Regulation of chemotaxis by the platelet-derived growth factor receptor- beta. Nature. 367,474476. Kwak, H. J., So, J. N., Lee, S. J., Kim, I., and Koh, G. Y. (1999). Angiopoietin-1 is an apoptosis survivd factor for endothelid cells. FEBS Lett. 448,249-253.

Ladbury, J. E., Lemmon, M. A., Zhou, M., Green, J., Botfield, M. C., and Schlessinger, J. ( 1995). Measurement of the binding of tyrosyl phosphopeptides to SH2 domains: a reappraisal. Proc. Natl. Acad. Sci. U.S. A. 92, 3 199-3203.

Lavan, B. E., Lane, W. S., and Lienhard, G. E. (1997). The 6û-kDa phosphotyrosine protein in insulin-treated adipocytes is a new member of the insulin receptor substrate family. J. Biol. Chem. 272, 1 1439-1 1443.

Lemmon, M. A., Ferguson, K. M., and Schlessinger, J. ( 1996). PH domains: diverse sequences with a common fold recruit signaling molecules to the ce11 surface. Cell85, 62 1-624.

Leveen, P., Pekny, M., Gebre-Medhin, S., Swolin, B., Larsson, E., and Betsholtz, C. ( 1994). Mice deficient for PDGF B show rend, cardiovascular, and hematological abnormalities. Genes. Dev. 8, 1875- 1887.

Li, N., Batzer, A., Daly, R., Yajnik, V., SkoInik, E., Chardin, P., Bar-Sagi, D., Margolis, B., and Schlessinger, J. ( 1993). Guanine-nucleotide-releasing factor hSos 1 binds to Grb2 and links receptor tyrosine kinases to Ras signalling. Nature. 363, 85-88.

Li, S. C., Songyang, Z., Vincent, S. J., Zwahlen, C., WiIey, S., Cantley, L., Kay, L. E., Forman-Kay, J., and Pawson, T. (1997). High-affinity binding of the Drosophila Numb phosphotyrosine-binding domain to peptides containing a Gly-Pro-(p)Tyr motif. Proc. NdAcad. Sei. U.S. A. 94, 7204-7209.

Li, S. C., Zwahlen, C., Vincent, S. J., McGlade, C. J., Kay, L. E., Pawson, T., and Forman-Kay, J. D. (1998). Structure of a Numb PTE! domain-peptide complex suggests a basis for diverse binding specificity. Nat. Struct. Biol. 5, 1075-1083.

Li. S. C., Gish, G., Yang, D., Coffey, A. J., Forman-Kay, J. D., Ernberg, I., Kay, L. E., and Pawson, T. (1999). Novel mode of ligand binding by the SH2 domain of the human XLP disease gene product SAP/SH2D 1 A. Curr. Biol. 9, 1355- 1362.

Li, W., Nishimura, R., Kashishian, A., Batzer, A. G., Kim, W. J., Cooper, J. A., and Schlessinger, J. (1994). A new function for a phosphotyrosine phosphatase: linking GRB2-SOSto a receptor tyrosine kinase. Mol. Cell Biol. 14,509-5 17.

Lin, P., Polverini, P., Dewhirst, M., Shan, S., Rao, P. S., and Peters, K. (1997). Inhibition of tumor angiogenesis using a soluble receptor establishes ri role for Tie2 in pathologie vascular growth. J. Clin. Invest. 100, 2072-2078. Lin, P., Buxton, J. A., Acheson, A., Radziejewski, C., Maisonpierre, P. C., Yancopoulos, G. D., Channon, K. M., Hale, L. P., Dewhirst, M. W., George, S. E., and Peters, K. G. ( 1998). Antiangiogenic gene therapy targeting the endothelium-specific receptor tyrosine kinase Tie2. Proc. Natl. Acad. Sci. CI. S. A. 95,8829-8834.

Lindahl, P., Johansson, B. R., Leveen, P., and Betsholtz, C. (1997). Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science. 277,242-245.

Lock, P., Casagranda, F., and Dunn, A. R. (1999). Independent SH2-binding sites mediate interaction of dok-related protein with RasGTPase-activating protein and Nck. J. Biol. Chem. 274,22775-22784.

Lowenstein, E. J., Ddy, R. J-. Batzer, A- G-,Li, W.. Margolis, B., Larnrners, R-,Ullrich, A., Skoinik, E. Y., Bar-Sagi, D., and Schlessinger, J. (1992). The SH2 and SH3 domain- containing protein GRB2 links receptor tyrosine kinases to ras signaling. Ce11 70,43 1- 443.

Macias, M. J., Musacchio, A., Ponstingl, H., Nilges, M., Saraste, M., and Oschkinat, H. ( 1994). Structure of the pleckstrin homology domain from beta-spectrin. Nature. 369, 675-677.

Maisonpierre, P. C., Goldfarb, M., Yancopoulos, G. D., and Gao, G. (1993). Distinct rat genes with related profiles of expression define a TIE receptor tyrosine kinase family. O~icogene.8, 163 1- 1637.

Maisonpierre, P. C., Suri, C., Jones, P. F., Bartunkova, S., Wiegand, S. J.. Radziejewski. C.. Compton, D., McClain, J., Aldrich, T. H., Papadopoulos, N., Daly, T. J.. Davis, S., Sato, T. N., and Yancopoulos, G. D. (1997). Angiopoietin-2, a natural antagonist for TieS that disrupts in vivo angiogenesis. Science. 277,5560-

Manser, J., Roonprapunt, C., and Margolis, B. (1997). C. elegans ce11 migration gene mig- 10 shares similarities with a family of SH2 domain proteins and acts ce11 nonautonomously in excretory canal development. Dev. Biol. 184, 150- 164.

Marengere, L. E., Songyang, Z., Gish, G. D., Schaller, M. D., Parsons, J. T., Stern, M. J., Cantley, L. C., and Pawson, T. (1994). SH2 domain specificity and activity modified by a single residue. Nature. 369,502-505.

Maroun, C. R., Holgado-Madmga, M., Royal, I., Naujokas, M. A., Fournier, T. M., Wong, A. J., and Park, M. (1999). The Gab 1 PH domain is required for localization of Gab 1 at sites of cell-ce1l contact and epithel ial morphogenesis downstream from the met receptor tyrosine kinase. Mol. Cell Biol. 19, 1784- 1799.

Marshall, C. J. (1995). Specificity of receptor tyrosine kinase signaling: transient versus sustained extracellular signal-regulated kinase activation. Cell. 80, 179- 185. Matsuda, M. and Kurata, T. ( 1996). Emerging components of the Crk oncogene product: the firs t identified adaptor protein. Cell Signal. 8, 335-340.

McGlade, J., Brunkhorst, B., Anderson, D., Mbarndu, G., Settleman, J., Dedhar, S., Rozakis-Adcock, M., Chen, L. B., and Pawson, T. ( 1993). The N-terminai region of GAP regulates cytoskeletal structure and ceIl adhesion. EMBO J. 12, 3073-308 1.

Meakin, S. O., MacDonald, J. I., Gryz, E. A., Kubu, C. J., and Verdi, J. M. (1999). The signaling adapter FRS-2 competes with Shc for binding to the nerve growth factor receptor TrkA. A mode1 for discriminating proliferation and differentiation. J. Biol. Cltern. 274, 986 1-9870.

Mezquita, J., Mezquita, B., Pau, M., and Mezquita, C. (1999). Characterization of a novel forrn of angiopoietin-2 (Ang-2B) and expression of VEGF and angiopoietin-2 during chicken testicular devetopment and regression. Biochem. Biophys. Res. Commun. 260, 492-498.

Millauer, B., Shawver, L. K., Plate, K. H., Risau, W., and Ullrich, A. (1994). Glioblastoma growth inhibited in vivo by a dominant-negative Flk- 1 mutant. Nature. 367. 576-579.

Millauer, B., Longhi, M. P., Plate, K. H., Shawver, L. K., Risau, W-. Ullrich, A., and Strawn, L. M. (1996). Dominant-negative inhibition of Flk-l suppresses the growth of many tumor types in vivo. Cancer Res. 56, 16 15- 1620.

Moran. M. F., Koch, C. A., Anderson, D., Ellis, C., England, L., Martin, G. S., and Pawson, T. (1990). Src homology region 2 domains direct protein-protein interactions in signal transduction. Proc. Narl. Acad. Sci U. S. A. 87, 8622-8626.

Moran, M. F., Polakis, P., McCorrnick, F., Pawson, T., and Ellis, C. ( 199 1). Protein- tyrosine kinases regulate the phosphorylation, protein interactions, subcelluIar distribution, and activity of p2 1 ras GTPase- activating protein. Mol. CelC Biol. 11, 1 804- 1812.

Mustonen, T. and Alitalo, K. (1995)- Endothelid receptor tyrosine kinases involved in angiogenesis. J. Ce11 Biol. 129, 895-898.

Myers, M. G.,Jr., Grammer, T. C., Brooks, J., Glasheen, E. M., Wang, L. M., Sun, X. J., Blenis, J., Pierce, J. H., and White, M. F. (1995). The pkckstrin homology domain in insulin receptor substrate- 1 sensitizes insulin signaling. J. Biol. Chem, 270, 1 17 15- 1 17 18.

NeIms, K., Snow, A. L., Hu-Li, J., and Paul, W. E. (1998). FFUP, a hematopoietic cell- specific rasGAP-interacting protein phosphorylated in response to cytokine stimulation. Itnmrrnity. 9, 13-24. Neufeld, G., Cohen, T., Gengrinovitch, S., and Poltorak, 2. ( 1999). Vascular endothelial growth factor (VEGF) and its receptors. FASEB. J. 13.9-22.

Noguchi, T., Matozaki, T., Inagaki, K., Tsuda, M., Fukunaga, K., Kitamura. Y., Kitamura, T., Shii, K., Yamanashi, Y., and Kasuga, M. (1999). Tyrosine phosphorylation of p62(Dok) induced by ce11 adhesion and insulin: possible role in cell migration. EMBO J. 18, 1748-1760.

O'Neill, T. J., Craparo, A., and Gustafson, T. A. (1994). Characterïzation of an interaction between insulin receptor substrate 1 and the insulin receptor by using the two- hybrid system. Mol. Ce12 Biol. 14,6433-6442.

Orlidge, A. and D'Amore, P. A. (1987). Inhibition of capiliary endothelial ceIl growth by pericytes and smooth muscle cells. J. Cell Biol. 105, 1455- 1462.

Ottinger, E. A., Botfield, M. C., and Shoelson, S. E. (1998). Tandem SH2 domains confer high specificity in tyrosine kinase signaling. J. Biol. Chern. 273, 729-735.

Pajusola, K., Aprelikova, O,, Armstrong, E., Morris, S., and Alitalo, K. (1993).Two human FLT4 receptor tyrosine kinase isoforms with distinct carboxy terminal tails are produced by alternative processing of primary transcripts. Oncogene. 8,293 1-2937.

Pandey, A., Lazar. D. F., Saltiel, A. R., and Dixit, V. M. (1994). Activation of the Eck receptor protein tyrosine kinase stimulates phosphatidylinositol 3-kinase activity. J. Biol. Clreni. 269, 30 154-30 1 57.

Papapetropoulos, A., Garcia-Cardena, G., Dengler, T. J., Maisonpierre, P. C., Yancopoulos, G. D., and Sessa, W. C. ( 1999). Direct actions of angiopoietin- 1 on human endothelium: evidence for network stabilization, cell survival, and interaction with other angiogenic growth factors. Lab. Invest. 79,2 13-223.

Pardanaud, L., Altmann, C., Kitos, P., Dieterien-Lievre, F., and Buck, C. A. (1987). Vasculogenesis in the early quail blastodisc as studied with a monoclonal antibody recognizing endothelial cells. Development. 100, 339-349.

Pardanaud, L., Yassine, F., and DieterIen-Lievre, F. ( 1989). Relationship between vascuIogenesis, angiogenesis and haemopoiesis during avian ontogeny. Developrnerzt. 105,473-485.

Partanen, J., Makela, T. P., Alitalo, R., Lehvaslaiho, H., and Alitalo, K. (1990). Putative tyrosine kinases expressed in K-562 human leukemia cells. Proc. Natl. Acad. Sci. U. S. A. 87, 89 1 3-89 17. Partanen, J., Armstrong, E., Makela, T. P., Korhonen, J., Sandberg, M., Renkonen, R., Knuutila, S., Huebner, K., and Alitalo, K. (1992). A novel endothelial tyrosine kinase with extracellular epidermal growth factor homology domains. Mol. Cell Biol. 12, 1698- 1 707.

Partanen, J. and Dumont, D. J. ( 1999). Functions of Tie 1 and Tie2 receptor tyrosine kinases in vascular development. Curr. Top. Microbiol. Immunol. 237, 159- 172.

Pascal, S. M., Singer, A. O., Gish, G., Yamazaki, T., Shoelson, S. E., Pawson, T., Kay, L. E., and Forman-Kay, J. D. (1994). Nuclear magnetic resonance structure of an SH2 domain of phospholipase C-gamma 1 complexed with a high affinity binding peptide. Cel177.46 1-472.

Patan, S., Munn, L. L., and Jain, R. K. (1 996). Intussusceptive microvascular growth in a human colon adenocarcinorna xenograft: a novel mechanism of tumor angiogenesis. Microvnsc. Res. 51, 260-272.

Patan, S. ( 1998). TE1 and TIE2 receptor tyrosine kinases inversely regulate embryonic angiogenesis by the mechanism of intussusceptive microvascular growth. Microvtzsc. Res. 56, 1-2 1.

Pawson, T. ( 1995). Protein modules and signalling networks. Nature. 373,573-580.

Pawson, T. and Scott, J. D. (1997). Signaling through scaffold, anchoring, and adaptor proteins. Science. 278, 2075-2080.

Pelicci, G., Lanfrancone, L., Grignani, F., McGlade, J., Cavallo, F., Forni, G., Nicoletti, I., Grignani, F., Pawson, T., and Pelicci, P. Ci. (1992). A novel transforming protein (SHC) with an SH2 domain is implicated in mitogenic signal transduction. Ce11 70. 93- 104.

Peters, K. G., Coogan, A., Berry, D., Marks, J., Iglehart, J. D., Kontos, C. D., Rao, P., Sankar, S., and Trogan, E. ( 1998). Expression of Tie2flek in breast tumour vasculature provides a new marker for evaluation of turnour angiogenesis. Br. J. Cancer 77. 5 1-56.

Ponzetto, C., BardeIli, A., Zhen, Z., Maina, F., dalla Zonca, P., Giordano, S., Graziani, A., Panayotou, G., and Comoglio, P. M. (1994). A multifunctional docking site mediates signaling and transformation by the hepatocyte growth factorlscatter factor receptor family . Cell77, 26 1-27 1 .

Poy, F., Yaffe, M. B., Sayos, J., Saxena, K., Morra, M., Sumegi, J., Cantley, L. C., Terhorst, C., and Eck, M. J. (1999). Crystal structures of the XLP protein SAP reveal a class of SH2 domains with extended, phosphotyrosine-independent sequence recognition. Mol. Ce11 4,555-56 1. Procopio, W. N., Pelavin, P. I., Lee, W. M., and Yeilding, N. M. (1999). Angiopoietin- 1 and -2 coiled coi1 domains mediate distinct homo- oligomerization patterns, but fibnnogen-like domains mediate ligand activity. J. Biol. Chem. 274,30 196-30201.

Puri, M. C., Rossant, J., Alitalo, K., Bernstein, A., and Partanen, J. (1995). The receptor tyrosine kinase TIE is required for integrity and survival of vascular endothelial cells. EMBO J. 14,5884-589 1.

Puri, M. C., Partanen, J., Rossant, J., and Bernstein. A. (1999). Interaction of the TEK and TEreceptor tyrosine kinases during cardiovascular development- Developrnent. 126. 4569-4580.

Ren, R., Mayer, B. J., Cicchetti, P., and Baltimore, D. (1993). Identification of a ten- amino acid proline-rich SH3 binding site. Science. 259, 1 157- 1 16 1.

Risau, W., Sariola, H., Zerwes, H. G., Sasse, J., Ekblorn, P., Kemler, R., and Doetschman, T. ( 1988). Vasculogenesis and angiogenesis in embryonic-stem-cell-derived embryoid bodies. Developrnent. 102,471-478.

Risau, W. (1995). Differentiation of endothelium. FASEB. J. 9,926-933.

Risau, W. and Flamme, 1. ( 1995). Vasculogenesis. Annu- Rev. Cell Dev. Biol. 11,73-91.

Risau, W. ( 1997). Mechanisms of angiogenesis. Nat~ire.386,67 1-674.

Rockow, S., Tang, J., Xiong, W., and Li, W. ( 1996). Nck inhibits NGF and basic FGF induced PC 12 ce11 differentiation via mitogen-acti vated protein kinase-independent pathway. Oncogene. 12. 235 1-2359.

Ross, S. A., Lienhard, G.E., and Lavan, B. E. (1998). Association of insulin receptor substrate 3 with SH2 domain-containing proteins in rat adipocytes, Biochern. Biophys. Res. Commun. 247,487-492.

Rozakis-Adcock, M., McGlade, J., Mbamalu, G., Pelicci, G., Daly, R., Li, W., Batzer, A., Thomas, S.. Bmgge, J., and Pelicci, P. G. (1992). Association of the Shc and GrbZSemS SHI-containing proteins is implicated in activation of the Ras pathway by tyrosine kinases. Nature. 360,689-692.

Rozakis-Adcock, M., Fernley, R., Wade, J., Pawson, T., and Bowtell, D. (1993). The SH2 and SH3 domains of mammalian Grb2 couple the EGF receptor to the Ras activator mSos 1. Nature. 363, 83-85.

Sadowski, I., Stone, J. C., and Pawson, T. (1986). A noncatalytic domain conserved among cytoplasmic protein-tyrosine kinases modifies the kinase function and transfonning activity of Fujinami sarcoma virus P 130gag-fps. Mol. Cell Biol. 6,4396- 4408.

Sariola, H., Ekblom, P., Lehtonen, E., and Saxen, L. (1983). Differentiation and vascularization of the metanephric kidney grafted on the chorioallantoic membrane. Dev. Biol. 96,427435.

Sato, A., Iwarna, A., Takakura, N., Nishio, H., Yancopoulos, G. D., and Suda, T. (1998). Characterization of TEK receptor tyrosine kinase and its ligands, Angiopoietins, in human hematopoietic progenitor cells. Int. Inzmunol. 10, 12 17- i 227.

Sato, T. N., Qin, Y., Kozak, C. A., and Audus, K. L. (1993). Tie-1 and tie-2 define another class of putative receptor tyrosine kinase genes expressed in earl y embryonic vascular system. Proc. Natl. Acad- Sci U. S. A. 90,9355-9358.

Sato, T. N., Tozawa, Y., Deutsch, U., Wolburg-Buchholz, K., Fujiwara, Y., Gendron- Maguire, M., Gridley, T., Wolburg, H., Risau, W., and Qin, Y. (1995). Distinct roles of the receptor tyrosine kinases Tie- 1 and Tie-2 in blood vesse1 formation. Nature. 376. 70- 74.

Sato. Y. and Rifkin, D. B. (1989). Inhibition of endothelial ce11 rnovernent by pericytes and srnooth muscle cells: activation of a latent transforrning growth factor-beta 1-like molecule by plasmin during co-culture. J. Cell Biol. 109, 309-3 15.

Schnurch, H. and Risau, W. (1993). Expression of tie-2, a member of a novel family of receptor tyrosine kinases, in the endothelial ce11 lineage. Developrnent. 119, 957-968.

Senger, D. R., Galli, S. J., Dvorak, A. M., Permzzi, C. A., Harvey V. S., and Dvorak, H. F. ( 1983). Turnor cells secrete a vascular pemeability factor that prornotes accumulation of ascites fluid. Science. 219,983-985.

Shalaby, F., Rossant, J., Yamaguchi, T. P., Gertsenstein, M., Wu, X. F., Breitman, M. L., and Schuh, A. C. ( 1995). Failure of blood-island formation and vasculogenesis in Flk- l- deficient mice. Nature. 376,62-66.

Shalaby, F., Ho, J., Stanford, W. L., Fischer, K. D., Schuh, A. C., Schwartz, L., Bernstein, A., and Rossant, J. (1997). A requirement for Flkl in primitive and definitive hematopoiesis and vasculogenesis. Ce11 89,98 1-990.

Shi, Q., Rafii, S., Wu, M. H., Wijelath, E. S., Yu, C., Ishida, A., Fujita, Y., Kothari, S., Mohle, R., Sauvage, L. R., Moore, M. A., Storb, R-F., and Hammond, W. P. (1998). Evidence for circutating bone marrow-derived endothelial cells. Blood. 92, 362-367.

Shi, Y. P., Mohapatra, G., Miller, J., Hanahan, D., Lander, E., Gold, P., Pinkel, D., and Gray, J. ( 1 997). FISH probes for mouse chromosome identification. Genomics. 45,42-47. Shibuya, M., Ito, N., and Claesson-Welsh, L. (1999). Structure and function of vascular endothelial growth factor receptor- 1 and -2. Curr. Top. Microbiol. Irnmunol. 237,59-83.

Shoelson, S. E., Sivaraja, M., Williams, K. P., Hu, P., Schlessinger, J., and Weiss, M- A. ( 1993). Specific phosphopeptide binding regulates a conformational change in the PI 3- kinase SH2 domain associated with enzyme activation. EMBO J. 12, 795-802.

Sholley, M. M., Ferguson, G. P., Seibel, H. R., Montour, J. L., and Wilson, J. D. (1984). Mechanisms of neovascularization. Vascular sprouting can occur without proliferation of endothelial cel1s. Lab. Invesr. 51, 624-634.

Siemeister, G., Schimer, M., Weindel, K., Reusch, P., Menrad, A,, Marrne, D., and Martiny-Baron, G. ( 1999). Two independent mechanisms essentiai for tumor angiogenesis: inhibition of human melanoma xenograft growth by interfering with either the vascular endothelial growth factor receptor pathway or the Tie-2 pathway. Cancer Res. 59,3185-3191.

Soker, S., Takashima, S., Miao, H. Q., Neufeld, G., and Kiagsbrun, M. (1998). Neuropilin- I is expressed by endotheliai and tumor cells as an isoform- specific receptor for vascuiar endotheiial growth factor. Cell92,735-745.

Songyang, Z., Shoelson, S. E., Chaudhuri, M., Gish, G., Pawson, T., Haser, W. G.. King, F.. Roberts, T., Ratnofsky, S., and Lechleider, R. J. (1993). SH2 domains recognize specific phosphopeptide sequences. Cell72,767-778.

Sosnowski, R. G., Feldman, S., and Feramisco, J. R. (1993). Interference with endogenous ras function inhibits cellular responses to wounding. J. Cell Biol. 121, 1 13- 1 19.

Stewart, P. A. and Wiiey, M. J. (198 1). Developing nervous tissue induces formation of blood-brain barrier characteristics in invading endothelial celIs: a study using quail-- chick transplantation chimeras. Dev. Biol. 84, 183- 192.

Stokoe, D., Stephens, L. R., Copeland, T., Gaffney, P. R., Reese, C. B., Painter, G. F., Holmes, A. B., McCormick, F., and Hawkins, P. T. (1997). Dual role of phosphatidylinositol-3,4,5-trisphosphate in the activation of protein kinase B. Science. 277,567-570.

Stratmann, A., Risau, W., and Plate, K. H. (1998). Ce11 type-specific expression of angiopoietin- 1 and angiopoietin-2 suggests a role in glioblastoma angiogenesis. Am. J. Patlzol. 153, 1459- 1466. Sun, X. J., Crimmins, D. L., Myers, M. G.,Jr., Miralpeix, M., and White, M. F. (1993). Pleiotropic insulin signals are engaged by multisite phosphorylation of IRS-1. Mol. Cell Biol. 13,7418-7428.

Suri, C., Jones, P. F., Patan, S., Bartunkova, S., Maisonpierre, P. C., Davis, S., Sato, T. N., and Yancopoulos, G. D. (1996). Requisite role of angiopoietin- 1, a ligand for the TE2 receptor, during embryonic angiogenesis. Cell87, 1 17 1- 1 180.

Suri. C., McClain, J., Thurston, G., McDonald, D. M., Zhou, H., Oldmixon, E. H., Sato, T. N., and Yancopoulos, G. D. (1998). Increased vascularization in mice overexpressing angiopoietin- 1. Science. 282.46847 1.

Taipale, J.. Makinen, T., Arighi, E., Kukk, E., Karkkainen, M., and Alitalo, K. ( 1999). Vascular endothelial growth factor receptor-3. Cwr. Top. Microbiol. Imrnunol. 237, 85- 96.

Takahashi, T., Kalka, C., Masuda, H., Chen, D.. Silver, M., Kearney, M., Magner, M., Isner, J. M., and Asahara, T. ( 1999). Ischemia- and cytokine-induced mobilization of bone marrow-derived endothelial progenitor cells for neovascularization. Nat. Med. 5, 434-43 8.

Takakura, N., Huang, X-L., Naruse, T., Hamaguchi, i., Dumont, D. J., Yancopoulos, G. D.. and Suda, T. (1998). Critical role of the TE2 endothelial ce11 receptor in the development of definitive hematopoiesis. Inimunity. 9,677-686.

Tanaka, S., Hattori, S., Kurata, T., Nagashima, K., Fukui, Y., Nakamura, S., and Matsuda, M. (1993). Both the SH2 and SH3 domains of human CRK protein are required for neurona1 differentiation of PC 12 cells. Mol. Cell Biol. 13,4409-44 15.

Tanaka, S.. Ouchi, T., and Hanafusa, H. ( 1997). Downstream of Crk adaptor signaling pathway: activation of Jun kinase by v-Crk through the guanine nucleotide exchange protein C3G. Proc. Nd.Acad. Sci U. S. A. 94,2356-2361.

Tang, J., Feng, G. S.. and Li, W. (1997). Induced direct binding of the adapter protein Nck to the GTPase- activating protein-associated protein p62 by epidermal growth factor. O~zcogerie.15, 1823- 1 832.

Teichert-Kuliszewska, K., Maisonpierre, P. C., Jones, N., Campbell, A.I.M., Master, 2.. Bendeck, M.P.,Alitalo, K.,Dumont, D. i., Yancopoulos, G. D., and Stewart, D. J. (2000) Biological actions of Angiopoietin- 1 and Angiopoietin-2 in a fibrin matrix model of angiogenesis. Submitted.

Thurston, G., Suri, C., Smith, K., McClain, J., Sato, T. N., Yancopoulos, G. D., and McDonald, D. M. (1999). Leakage-resistant blood vessels in mice transgenically overexpressing angiopoietin- 1. Science. 286,25 1 1-25 14. Trahey, M., Wong, G., Halenbeck, R., Rubinfetd, B., Martin, G. A., Ladner, M., Long, C. M., Crosier, W. J., Watt, K., and Kohs, K. (1988). Molecular cloning of two types of GAP complementary DNA from human placenta. Science. 242, 1697- 1700.

Valenzuela, D. M., Griffiths, J. A., Rojas, J., Aldrich, T. H., Jones, P. F., Zhou, H., McClain, J., Copeland, N. G., Gilbert, D. J., Jenkins, N. A., Huang, T., Papadopoulos, N., Maisonpierre, P. C., Davis, S., and Yancopoulos, G. D. (1999). Angiopoietins 3 and 4: diverging gene counterparts in mice and humans. Proc. Natl. Acad. Sci U.S. A. 96, 1904- 1909. van der Geer, P., Hunter, T., and Lindberg, R. A. (1994). Receptor protein-tyrosine kinases and their signal transduction pathways. Annu. Rev. Cell Biol. 10, 25 1-337. van der Geer, P., Wiley, S., Gish, G. D., Lai, V. K.. Stephens, R., White, M. F., Kaplan, D., and Pawson, T. (1996). Identification of residues that control specific binding of the Shc phosphotyrosine-binding domain to phosphotyrosine sites. Proc. Narl. Acad. Sci. U. S. A. 93,963-968. van der Geer, P., Henkemeyer, M., Jacks, T., and Pawson, T. ( 1997). Aberrant Ras reguIation and reduced pl90 tyrosine phosphoryiation in ceils Iacking pl2O-Gap. Mol. Ceil Biol. 17, I 840- 1 847.

Vecchi, A., Garlanda, C., Larnpugnani, M. G., Resnati. LM.,Matteucci, C., Stoppacciaro, A.. Schnurch, H., Risau, W., Ruco, L., and Mantovani, A. (1994). Monoclonal antibodies specific for endothelia1 cells of mouse blood vessels. Their application in the identification of adult and embryonic endothelium. Eur. J. Cell Biol. 63, 247-254.

Vikkula, M., Boon, L. M., Carraway, K. L., Calvert. J. T., Diamonti, A. J-, Goumnerov, B.. Pasyk, K. A., Marchuk, D. A., Warman, M. L., Cantley, L. C., Mulliken, J. B., and Olsen, B. R. (1996). Vascular dysmorphogenesis caused by an activating mutation in the receptor tyrosine kinase TIE2. Ce11 87, 1 18 1- 1 190.

Voliovitch, H., Schindler, D. G., Hadari, Y. R., Taylor, S. I., AcciIi, D., and Zick, Y. ( 1995). Tyrosine phosphorylation of insulin receptor substrate- 1 in vivo depends upon the presence of its pleckstrin homology region. J. Biol. Chem. 270, 18083- 18087.

Waksman, G., Kominos, D., Robertson, S. C., Pant, N., Baltimore, D., Birge, R. B., Cowburn, D., Hanafusa, H., Mayer, B. J., and Overduin, M. (1992). CrystaI structure of the phosphotyrosine recognition domain SH2 of v- src complexed with tyrosine- phosphory lated peptides. Nature. 358,646-653.

Weidner, K. M., Di Cesare, S., Sachs, M., Brinkmann, V., Behrens, J., and Birchmeier, W. ( 1996). Interaction between Gabl and the c-Met receptor tyrosine kinase is responsible for epithelial morphogenesis. Nature. 384, 173- 176. Welharn, M. J., Dechert, U., Leslie, K. B., Jirik, F., and Schrader, J. W. (1994). Interleukin (IL)-3 and granulocyte/macrophage colony-stimulating factor, but not IL-4, i nduce tyrosine phosphory lation, activation, and association of SHPTP2 with Grb2 and p hosphatidylinositol 3'-kinase. J. Biol. Chem. 269, 23764-23768.

Wilting, J. and Christ, B. (1996). Embryonic angiogenesis: a review. Nf~tf(nvissenschafien.83, 153-164.

Witzenbichler, B., Maisonpierre, P. C., Jones, P., Yancopoulos, G. D., and Isner, J. M. ( 1998). Chernotactic properties of angiopoietin- 1 and -2, ligands for the endothelial- specific receptor tyrosine kinase Tie2. J. Biol. Chem. 273, 185 14- 18521.

Wolf, G., Trub, T., Ottinger, E., Groninga, L., Lynch, A., White, M. F., Miyazaki, M., Lee. J., and Shoelson, S. E. (1995). PTB domains of IRS-1 and Shc have distinct but overlapping binding specificities. J. Biol. Chem. 270, 27407-274 10.

Wong, A. L., Haroon, 2. A., Werner, S., Dewhirst, M. W., Greenberg, C. S., and Peters, K. G. ( 1997). Tie2 expression and phosphorylation in angiogenic and quiescent adult tissues. Circ. Res. 81, 567-574.

Xu, P., Jacobs, A. R., and Taylor, S. 1. (1999). Interaction of insulin receptor substrate 3 with insulin receptor, insulin receptor-related receptor, insulin-like growth factor- 1 receptor, and downstream signaling proteins. J. Biul. Chern- 274, 1 5262-1 5270.

Yarnanashi, Y. and Baltimore, D. (1997). Identification of the Abl- and rasGAP- associated 62 kDa protein as a docking protein, Dok. Cell88,205-2 1 1.

Yamanashi, Y., Tamura, TT.,Kanamori, T., Yamane, H., Nariuchi, H., Yamamoto, T., and Baltimore, D. (2000). Role of the rasGAP-associated docking protein p62(dok) in negative regulation of B ce11 receptor-mediated signaling. Genes. Dev. 2000. Jan. /;IJ(l):If-16. 14, 11-16.

Yenush, L., Makati, K. J., Smith-Hall, J., Ishibashi, O., Myers, M. G.,Jr., and White, M. F. (1996). The pleckstrin homology domain is the principal link between the insuiin receptor and IRS- 1. J. Biol. Chem. 271,24300-24306.

Yenush, L. and White, M. F. (1997). The IRS-signalling system during insulin and cytokine action. Bioassays. 19,491-500.

Yokote, K., Margolis, B., Heldin, C. H., and Claesson-WeIsh, L. (1996). Grb7 is a downstrearn signaling component of platelet-derived growth factor alpha- and beta- receptors. J. Biol. Chem. 271,30942-30949. Yu, D. H.,Qu, C. K., Henegariu, O., Lu, X., and Feng, G. S. (1998). Protein-tyrosine phosphatase Shp-2 regulates ce11 spreading, migration, and focal adhesion. J. Biol. Chem. 273,21125-21131.

Yuan, H. T., Suri, C., Yancopoulos, G. D., and Woolf, A. S. (1999). Expression of angiopoietin- 1, angiopoietin-2, and the Tie-2 receptor tyrosine kinase during mouse kidney maturation. J. Am. Soc. Nephrol. 10, 1722- 1736.

Zagzag, D., Hooper, A., Friedlander, D. R., Chan, W., Holash, J., Wiegand, S. J.. Yancopoulos, G. D., and Grumet, M. (1999). In situ expression of angiopoietins in astrocytomas identifies angiopoietin-2 as an early marker of tumor angiogenesis. Exp. Newol. 159,391 -400.

Zhao, C., Yu, D. H., Shen, R., and Feng, G. S. (1999). Gab2, a new pleckstrin homology dornain-containing adapter protein, acts to uncouple signaling from ERK kinase to Elk- 1. J. Biol. Chern. 274, 19649-19654.

Zhou, M. M., Ravichandran, K. S., Olejniczak, E. F., Petros, A. M., Meadows, R. P., Sattler, M., Hahn. J. E., Wade, W. S., Burakoff, S. J., and Fesik, S. W. (1995). Structure and ligand recognition of the phosphotyrosine binding domain of Shc. Nature. 378, 584- 592.

Zhou, M. M., Huang. B., Olejniczak, E. T., Meadows, R. P., Shuker, S. B., Miyazaki, M., Trub, T., Shoelson, S. E., and Fesik, S. W. (1 996). Structural basis for IL-4 receptor phosphopeptide recognition by the IRS- 1 PTB domain. Nar. Struct. Biol. 3, 388-393.

Zhou, S. and Cantley, L. C. (1995). Recognition and specificity in protein tyrosine kinase-mediated signailing. Trends- Biochern. Sei. 20,470-475.

Ziegler. S. F., Bird, T. A., Schneringer, J. A., Schooley, K. A., and Baum, P. R. ( 1993). Molecular cloning and characterization of a novei receptor protein tyrosine kinase from human placenta, Oncogerze. 8,663-670.